WO2011107994A1

WO2011107994A1 - Methods of measuring protein stability

Info

Publication number: WO2011107994A1
Application number: PCT/IL2011/000213
Authority: WO
Inventors: Uri Alon; Eran Eden; Naama Geva-Zatorsky
Original assignee: Yeda Research And Development Co. Ltd.
Priority date: 2010-03-04
Filing date: 2011-03-03
Publication date: 2011-09-09

Abstract

A method of determining a degradation rate of a fluorescently labeled polypeptide of interest in a population of cells is provided. The method comprises: (a) obtaining two identical populations of cells; (b) determining fluorescence levels of the fluorescently labeled polypeptide of interest in a first of the at least two populations of cells at at least two different points in time; (c) at least partially reducing a fluorescence of the fluorescently labeled polypeptide of interest in a second of the at least two populations of cells; (d) determining fluorescence levels of the reduced fluorescently labeled polypeptide of interest in the second of said at least two populations of cells at said at least two different points in time; (e) comparing fluorescence levels in the first and the second populations; and (f) based on thee comparing, determining a degradation rate of the fluorescently labeled polypeptide of interest.

Description

METHODS OF MEASURING PROTEIN STABILITY

FIELD AND BACKGROUND OF THE INVENTION

The present invention, in some embodiments thereof, relates to a method of measuring protein stability inside a cell and system capable of same.

Protein molecules are continuously synthesized and degraded in all living organisms. The concentration of individual cellular proteins is determined by a balance between the rates of synthesis and degradation, which in turn are controlled by a series of regulated biochemical mechanisms. Differences in the rates of protein synthesis and breakdown result in cellular and tissue atrophy (loss of proteins from cells) and hypertrophy (increase in protein content of cells).

The dynamics of proteins are controlled at several levels, including transcription translation and degradation. In bacteria and single celled eukaryotes, transcription is probably the most important determinant of protein levels [1-7]. Although translation and degradation control are important in these organisms, much information can be gleaned from transcription assays. In contrast, in organisms such as mammals, the correlation between mRNA levels and protein levels is less clear. Snapshots of mRNA and proteins at a single time point suggests low correlation across different genes in humans and mouse [8-13]. Current technology makes it difficult to go beyond snapshots, especially due to the challenge of accurate proteomics.

Analysis of post-transcription control in mammalian cells is challenging. In the classic cell-cycle experiment that proved the existence of cyclins, Hunt et al., [Evans, T. et al. Cell 33, 389-396 (1983)] labeled sea urchin eggs with radioactive methionine and used pulse-chase analysis to show periodic destruction of a mysterious band at distinct time points after fertilization. Similar pulse-chase experiments with and without protein- synthesis inhibitors are now commonly used to measure the stability of individual proteins, but these assays are not amenable to high-throughput characterization of mammalian cells. Mass spectrometry and stable isotope-labeling approaches have been used to characterize protein turnover on a larger scale but have suffered from insufficient dynamic range and the confounding effects of cell growth and ongoing protein synthesis [19]. In yeast, high-throughput studies have defined half-lives for >3,750 proteins using a comprehensive library of genetically tagged strains and Western blot analysis [Belle, A. et al. Proc. Natl. Acad. Sci. USA 103, 13004-13009]. However, the effects of epitope tags on transcription and translation remain as confounding factors, and no such integrated expression library exists for mammalian cells. A limitation common to all these methods is that protein stability is measured at the population level rather than in individual cells.

Yen et al [20] teach a method for determining global protein stability (GPS) profiling that couples flow cytometry with microarray technology. The method comprises delivery of a retroviral, dual-fluorescent reporter system that measures the abundance of target proteins while simultaneously controlling for experimental variability and cell-to-cell differences in transcription and translation.

SUMMARY OF THE INVENTION

According to an aspect of some embodiments of the present invention there is provided a method of determining a degradation rate of a fluorescently labeled polypeptide of interest in a population of cells, the method comprising:

(a) obtaining two identical populations of cells;

(b) determining fluorescence levels of the fluorescently labeled polypeptide of interest in a first of the at least two populations of cells at at least two different points in time;

(c) at least partially reducing a fluorescence of the fluorescently labeled polypeptide of interest in a second of the at least two populations of cells;

(d) determining fluorescence levels of the reduced fluorescently labeled polypeptide of interest in the second of the at least two population of cells at the at least two different points in time;

(e) comparing fluorescence levels in the first and the second populations; and

(f) based on the comparing, determining a degradation rate of the fluorescently labeled polypeptide of interest.

According to some embodiments of the invention, step (f) is effected by calculating the ratio of the (natural log (In) of the difference in fluorescence of the polypeptide of interest in the first population of cells and fluorescence of the polypeptide of interest in the second population of cells at a first point in time): (In of the difference in fluorescence of the polypeptide of interest in the first population of cells and fluorescence of the polypeptide of interest in the second population of cells at a second point in time) divided by the difference in time between the two points in time.

According to some embodiments of the invention, step (f) is effected by determining a slope of a graph, wherein a y axis of the graph indicates the In (P(t) - Pv(t)) and an x axis of the graph indicates time, wherein P(t) is the amount of fluorescence of the polypeptide of interest at point t in time in the first population of cells and (Pv(t) is the amount of fluorescence of the polypeptide of interest at point t in time in the second population of cells.

According to some embodiments of the invention, step (c) is effected with a mercury fluorescent lamp.

According to some embodiments of the invention, step (c) is effected for 1-8 minutes.

According to some embodiments of the invention, the population of cells comprises an additional fluorescently labeled polypeptide, the additional fluorescently labeled polypeptide having a higher nuclear: cytoplasm expression ratio and the additional fluorescently labeled polypeptide being distinguishable from the fluorescently labeled polypeptide of interest.

According to some embodiments of the invention, the polypeptide of interest is transcribed from its native location in the population of cells.

According to an aspect of some embodiments of the present invention there is provided a method of analyzing the effect of an agent on a degradation rate of a polypeptide of interest in a population of cells, the method comprising:

(a) contacting a population of cells with the agent;

(b) determining the degradation rate of the fluorescently labeled polypeptide of interest in the population of cells according to the method of the present invention; and

(c) comparing the degradation rate obtained in step (b) with a degradation rate of the polypeptide in an absence of the agent, thereby analyzing the effect of the agent on the degradation rate of the polypeptide of interest.

According to an aspect of some embodiments of the present invention there is provided a system for determining a degradation rate of a fluorescently labeled polypeptide of interest in a cell comprising a processing unit, the processing unit executing a software application configured for converting change in fluorescence levels over a period of time to a degradation rate.

According to some embodiments of the invention, the degradation rate is determined according to the method of the present invention.

According to some embodiments of the invention, the system further comprises an imaging system.

According to some embodiments of the invention, the imaging system comprises an image capture apparatus.

According to some embodiments of the invention, the image capture apparatus is capable of capturing an image of the fluorescently labeled polypeptide over the period of time.

According to some embodiments of the invention, the imaging system further comprises a data processor for calculating a fluorescence level for each picture element of the image.

According to some embodiments of the invention, the image capture apparatus is a fluorescent image capture apparatus.

Unless otherwise defined, all technical and/or scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the invention pertains. Although methods and materials similar or equivalent to those described herein can be used in the practice or testing of embodiments of the invention, exemplary methods and/or materials are described below. In case of conflict, the patent specification, including definitions, will control. In addition, the materials, methods, and examples are illustrative only and are not intended to be necessarily limiting.

Implementation of the method and/or system of embodiments of the invention can involve performing or completing selected tasks manually, automatically, or a combination thereof. Moreover, according to actual instrumentation and equipment of embodiments of the method and/or system of the invention, several selected tasks could be implemented by hardware, by software or by firmware or by a combination thereof using an operating system.

For example, hardware for performing selected tasks according to embodiments of the invention could be implemented as a chip or a circuit. As software, selected tasks according to embodiments of the invention could be implemented as a plurality of software instructions being executed by a computer using any suitable operating system. In an exemplary embodiment of the invention, one or more tasks according to exemplary embodiments of method and/or system as described herein are performed by a data processor, such as a computing platform for executing a plurality of instructions. Optionally, the data processor includes a volatile memory for storing instructions and/or data and/or a non -volatile storage, for example, a magnetic hard-disk and/or removable media, for storing instructions and/or data. Optionally, a network connection is provided as well. A display and/or a user input device such as a keyboard' or mouse are optionally provided as well.

BRIEF DESCRIPTION OF THE DRAWINGS

Some embodiments of the invention are herein described, by way of example only, with reference to the accompanying images. With specific reference now to the drawings in detail, it is stressed that the particulars shown are by way of example and for purposes of illustrative discussion of embodiments of the invention. In this regard, the description taken with the drawings makes apparent to those skilled in the art how embodiments of the invention may be practiced.

In the drawings:

FIGs. 1A-K are graphs and spectra illustrating that anti-cancer drug, CPT induces waves of time-dependant mRNA changes across different groups of functionally related genes. GO enrichment analysis was applied in order to indentify groups of genes sharing the same function, process or localization that also respond similarly to the drug. To this end, genes were ranked according to their level of expression at each time point and given as input to the Gorilla software which identifies and visualizes the enriched GO terms (for details see Method). Many cellular processes and functions were identified that responded in concert to the drug in a statistically significant manner (P < 10^"4 after multiple hypothesis correction) thus revealing different affects of the drug. In the figure, each enriched GO-term is depicted by three subplots. The top subplot is a 'GO-Gram' that shows the enrichment log P-value of the GO-term as a function of time with positive and negative values indicating whether the enrichment was attained in over-expressed or under expressed genes (values are truncated at + 20). The middle and bottom subplots show the average and the individual mRNA dynamics of all the genes associated with the GO-term. (A) Histones and genes responsible for nucleosome assembly are dramatically down regulated immediately upon drug addition (P<10^"60) and then gradually recover to about half of the initial levels. (B) Transcription factors that control expression prior to drug addition are shut-down immediately upon drug addition followed by rise of an alternative set of transcription factors after 1.5 hours. (C) Surprisingly, stress related genes and DNA repair genes are significantly down-regulated 1.5 - 30 hours after the drug followed by a recovery. (D) Cell division related genes and DNA replication are down regulation ~5 hours after drug addition. This is in line with observations that the cells stop dividing after the drug and reach a complete halt after ~5 hours. Interestingly, an over-expression spike is observed of a few genes regulating cell proliferation just before the decrease in mitosis related genes begins. (E) RNA processing declines in response to the drug. (F) The drug induces a spike of kinase regulating genes, followed by a decrease of serine/threonine kinases. (G) MHC-I related genes exhibit a time- dependant increase in response to the drug. This finding is consistent with the role of MHC-I in other stress responses [1]. (H - K) Additional significant response include: increased expression of energy and metabolite generation related genes, down regulation of cytoskeletal related genes that is correlated in time with our measured decrease in cell motility, a transient decrease and then recovery of ubiquitine dependant catabolic processes and an increase in lysosomal genes.

FIG. 2 is a bar graph illustrating mRNA-protein temporal correlation distribution in real and shuffled data. The correlation distribution of mRNA and protein temporal drug responses is indicated by the gray bars. This was compared to the correlation distribution generated by pairing of the protein profiles with mRNA profiles that were randomly picked (blue bar). Note that the actual profiles were not shuffled, only the pairing. The empirical distribution showed about 15 % more positive correlations than in the random pairing distribution (KS P < 10^"6 ).

FIG. 3 are photographs and readout illustrating that the co-dynamics of 540 proteins and their respective mRNAs transcripts were measured for 48 hours after the addition of the anti-cancer drug, CPT. Levels of 540 endogenously YFP tagged proteins were measured for 48 hours (images were taken every 20 minutes) in human cancer cells using time laps fluorescence microscopy. Custom made software was used for automatic video analysis including automatic cell tracking, phenotype detection and protein information retrieval. Additionally, the mRNA levels of all human transcripts were measured under the same conditions using Affymetrix GeneChip Human Gene 1.0 ST arrays at time points 0, 1.5, 3, 6, 9, 12, 23, 33, 48 hours after drug addition. Proteins were coupled to their respective mRNA resulting in a dataset of 540 protein-mRNA co- dynamic profiles.

FIG. 4 is a workflow of PhenoTrack - a system for automatic cell tracking and phenotype retrieval from time-laps movies. PhenoTrack receives as input 3 videos: phase contrast, protein tagged with YFP and cherry tag for image analysis. The output includes information on the individual cell including: protein levels, cell sizes, velocities, mitosis rate, time of death (in case cell dies) etc. Information is stored in a database and can be visualized.

FIGs. 5A-B are graphs illustrating day to day mRNA level reproducibility error is better than 15%. (A) the CV (std/mean) of all genes over all time points were computed using at least three independent experiments . The attained CV distribution has a median of less than 15% indicating that changes in the dynamics that are larger than 20 % can be detected. (B) Examples of mRNA fold change and the corresponding standard error sfcd/ n ) over day to day repetitions. It can be seen that the measured mRNA dynamics is substantially larger than the experimental error.

FIG. 6 are graphs illustrating mRNA measurements obtained by microarrays were validated using qPCR. The drug response of 10 genes were measured using qPCR at times 0, 2, 9, 23 33 and 48 hours after drug addition (with at least three biological replicates at each time point) and compared to the measurements obtained using microarrays. The mRNA levels were normalized to levels before drug addition (t = 0) with blue and red lines indicate the qPCR and microarray mRNA measurements respectively. The control house keeping gene GAPDH, retained constant levels and hence there was no need to use it for normalization. The profiles obtained by the qPCR and the microarray are highly similar (median correlation r = 0.83).

FIG. 7 are graphs illustrating that mRNA and protein pairs fall into major co- dynamic clusters. The temporal responses of mRNA and protein pairs were concatenated and clustered using k-means algorithm with 4 seeds. A few distinct general patterns were observed: (i) pairs in which the mRNA level is increases and the protein levels decrease in an anti correlated fashion; (ii) pairs in which mRNAs decreases and protein is constant or increases, (iii) degrading proteins whose mRNA first degrades and then starts to build up and (iv) correlated mRNA and proteins both of which have a short period of increase followed by a decrease.

FIGs. 8A-I are graphs illustrating that the response of most of the transcriptome and proteome is dis-correlated in time. (A) The distribution of the temporal correlation of 540 protein-mRNA pairs in response to a drug is shown. Approximately 30 % of the pairs showed a strong positive temporal correlation (p > 0.6); 15% were negatively correlated (p < -0.6) and the remaining 55% showed a weak or no correlation. (B) Examples of protein-mRNA pairs showing different type of behaviors: positive correlation (MYH9), time shift positive correlation (CKS2), no-correlations (EIF4A1, HDAC) and negative correlations (NCBP2, ENOl). (C - F) Functional groups of genes show a systematic dis-correlation or correlation between their proteins and mRNAs. (C, D) Ribosomal proteins and their mRNAs are dis-correlated in a systematic manner. Each row corresponds to a different ribosomal protein and its respective mRNA with yellow/red and red/green indicating high and low levels respectively. Protein and mRNA levels of each pair were linearly scaled between -1 and 1. While the ribosomal proteins decay in response to the drug their respective mRNAs increase in the first 12 hours and then begin to decrease after 24 hours. (E) Average response of ribosomal proteins and mRNAs (indicated by blue and red lines respectively) is weakly correlated (r = 0.37). Protein averages were normalized to t = 0 hours. (F, G) The dynamics of most cytoskeletal associated proteins are synchronized with their mRNA dynamics. (H) Average response of cytoskeletal associated proteins and their mRNAs is highly correlated (r = 0. 86). (I) mRNA levels do not correlate well with relative protein levels across different genes. Each dot represents a pair of mRNA and protein measurements at a single time point. Pearson correlation across all mRNAs and protein abundances was found to be weak (r = 0.13) yet significant (P < 10^"9).

FIGs. 9A-C are graphs and drawings illustrating the difference between bleached and non-bleached protein fluorescence decays in time at a rate that is exponential in the protein degradation rate. (A) The bleach-chase protocol is illustrated: protein fluorescent levels are measured at different time -points in bleached and non- bleached cells. The levels of the non-fluorescent proteins in the bleached cells are indirectly measured by subtracting the two profiles. Their decay in time reveals the protein degradation rate and half life. (B) The top panel illustrates three proteins that respond to the same stimulus given at t = 0 without bleaching (blue lines) and with bleaching (red broken lines). The log rate at which the bleached and un-bleached protein levels converge in each of the examples is indicated in the lower panel. The slope reveals the degradation rate according to the following equation: ln (P(t)— P_v(t)) = ίη(Ρ₀— ¾,}— a t, where ^p and ^pv are the protein levels with and without bleaching and a is the degradation rate. Hence, a bleached and non- bleached response that converges rapidly (upper left box) indicates rapid degradation where as parallel dynamics (upper right box) indicate no degradation at all. Note that the rate at which the dynamics converge does not depend on the production rate. (C) Degradation rate measurements using bleach-chase are invariant to the amount of bleaching. The CPT drug response of the ribosomal protein RPS3A was monitored after 0, 2 and 4 minutes of bleaching for 20 hours (indicated by blue, red and green lines in the upper box). The log convergence between 0 and 2 as well as 0 and 4 minutes of bleaching is shown in the lower box (green and red lines are the respective linear fits) both of which yield similar slopes and hence similar degradations (0.08 and 0.09 1 / hour).

FIGs. 10A-D are graphs illustrating that protein half-lives increase upon drug addition: the larger the half life, the larger the increase. (A) A comparison of the degradation rates before and after drug addition revealed that most proteins either decrease or retain the same level of degradation. (B) Examples of protein degradation after drug addition. Top panel shows four proteins (RPS3A, EEF2, DDX5 and LMNA) as they respond to the drug with and without bleaching indicated by pairs of red and blue lines respectively. Fluorescent levels were normalized to the unbleached levels at t = 0. Bottom panel shows the log difference between the bleached and unbleached responses as a function of time. The slope of the linear fit, a, is the protein degradation rate in 1/hours units. (C) The half-lives of most proteins after drug addition either increase or remain constant compared to their levels before drug addition. The increase in half-lives is stronger the longer the pre-drug lifetime of the protein is (as illustrated in the inner box). Blue line is the predicted shift in half-lives due to growth arrest. It is obtained by setting T_cc to 20 hours (the cell-cycle duration in the present cells), k to 0 (since drug halts cell-cycle) and applying Eq. 6. The predicted behavior captures more than 40% of the variance in the data. (D) Increase in protein stability due to growth arrest is predicted to be inversely proportional to the cell cycle duration. Cell cycle duration, T_cc, is set to 4 hours, 1 day, 1 week and infinity (i.e. no division) and the corresponding half-life behaviors before and after growth arrest are plotted by applying Eq. 6. The model predicts that the faster cells divide prior to drug addition, the stronger the global increase in half -lives.

FIGs. 11A-E are graphs illustrating how protein translation rates decrease in response to the drug. (A-B) Average translation rate and individual protein translation rates are shown. Translation rates were normalized to levels prior to drug addition. Most proteins show a decrease in translation in response to the drug, with a mean response that is about half of the initial level. (C) Average ribosomal protein drug response shows a decrease that is synchronized with timing of translation decay. (D) Changes in half-life and translation rates before and 48 hours after drug addition are shown for each of the proteins (indicated by the trajectory connecting each pair of blue and red dots). The drug induces two opposing effects: a global increase in protein half- lives (indicated by the global shift from left to right) and simultaneous decrease in translation rates (indicated by the global shift from top to bottom). The balance between these counter effects varies considerably across different proteins and strongly affects protein dynamics. (E) Examples of four protein dynamics with dis-correlated and even anti-correlated mRNA responses. The proteins exhibit a decrease in translation and an increase in half-lives the balance of which determines protein behavior, thus resolving the dis-correlation between the mRNA and protein profiles in these proteins.

FIG. 12 are graphs illustrating an integrated view of protein, mRNA, half-life and translation rate dynamics in response to the drug. The time dependant changes 0 - 48 hours after drug addition in four different dimension are shown (i) protein fluorescence dynamics normalized to t = 0; (ii) mRNA dynamics normalized to t = 0; (iii) changes in protein half-lives measured in hours and (iv) changes in translation rates normalized to t = 0. Proteins are ordered according to increasing levels of post drug half-lives. It can be seen that most proteins show a time dependant decrease in their translation rate and an increase in half-life. Proteins that degraded tend to have a small or no increase in their in half-lives where as proteins that retained a constant level or increased tend to show larger post-drug half-lives (e.g. ENOl, NASP, H2AFV and LMNA) indicating that half-levels after drug addition play a major role in dictating protein dynamics.

FIG. 13 are graphs of protein profiles in response to the translation inhibitor CHX. The dynamic profiles of 1 1 proteins were measured in response to ΙΟμΜ of the translation inhibitor CHX. Colored lines and back line indicate the response of individual cells and the average response respectively. All values were normalized to the average level at t = 0.

FIG. 14 is a graph illustrating global increase in half-lives due to intra-cellular degradation inhibition does not fit the data. Protein half-lives before and after drug addition are shown. Predicted increase in half-lives due to of intra-cellular degradation inhibition is shown as well, where different k's dictate the degree of inhibition according to Eq. S9. It can be seen that although this model predicts a global increase in protein half-lives, it does not account for the observation that stable proteins become increasingly more stable. On the contrary, stable proteins are expected to be the least affected, since their half-lives are largely determined by cell-division rather than intracellular degradation.

FIGs. 15A-B are examples of autoradiographs pulse-chase results for MYH9

(Figure 15B) and DDX5 (Figure 15A). The H1299 cells expressing the YFP tagged protein of interest were pulse-labeled with [35S] methionine/cysteine followed by cold chase as described in the Methods. At the times indicated, the cells were collected and whole cell lysates were prepared. The proteins of interest were recovered from 200 μg of cell lysates by immunoprecipitation (IP). Anti-GFP antibody was used to precipitate the YFP tagged proteins; protein-specific Abs were used to precipitate both the YFP tagged and the wild type forms of the proteins; non-specific rabbit or mouse IgG was used as a negative control. The immunoprecipitated proteins were resolved on SDS- PAGE and visualized by autoradiography (the left panels). Parallel gels were blotted onto nitrocellulose membrane and probed with the indicated antibodies (the right panels) to ensure correct identification of the protein bands of interest. About 20-40 g of cell lysates (10-20% of the amount used for IP) were loaded on the gels as positive controls (Input). Each experiment was done with at least two biological replicates. To determine protein removal rates, intensities of the relevant protein bands were quantified from the autoradiographs using a custom Matlab software. Levels of the radioactive labeled protein were plotted on a semi-logarithmic scale. The slope of the linear fit is the removal rate.

FIGs. 16A-B are graphs illustrating that removal rate measurements are invariant to the amount of bleaching. The CPT response of RPS3A was measured without bleaching (blue dots) and after two and four minutes of bleaching (red and green dots respectively). The convergence rate in both cases yielded similar removal rates (0.08 ± 0.01 and 0.09 ± 0.01 1 / hour).

FIGs. 17A-D are graphs illustrating that the balance between degradation and dilution under normal growth varies widely between proteins. Distribution of 100 protein removal rates (A) and half-lives (B) under normal growth, a ranges between 0.03 and 0.82 with average 0.1±0.09 (1/h). (C) The balance between degradation and dilution of 100 proteins. (D) Proteins with similar functions or localizations tend to share similar half-lives.

FIGs.l8A-B are graphs illustrating protein half-lives at t=0-24 and t=24-48 hours after addition remain roughly constant or increase. (A, B) The half-lives of 10 proteins at t = 0 - 24 and t = 24 - 48 hours after drug addition are compared. Most half- lives remained approximately constant or slightly increased. CALM2 and K-a-1 showed a substantial increase in their half-lives. Blue line indicates the predicted increase due to growth arrest.

FIGs. 19A-G are graphs illustrating that protein half-lives increase in response to stress: the longer the half-life, the larger the increase. (A) Half-lives in CPT compared to normal growth. Blue line is the predicted half-lives due to growth arrest (Eq. 5), with T_cc =22.5h (measured cell-cycle duration) and k₂=0 (growth arrest). White dots indicate proteins that significantly deviate from line. (B) A model with reduced degradation (Eq. 2-3) does not account for the observed half-lives, whereas a model with reduced dilution (Eq. 4-5) does (C). (D-G) Changes in protein half-lives, comparing normal growth to different stresses, are captured by the reduced dilution model (blue lines), with k₂ equal to the measured ratio between post and pre-stress growth rates. FIG. 20 is a graph illustrating that the expected increase in half-lives due to degradation inhibition does not account for the observed response to the drug CPT. Reducing degradation is expected to increase protein half-lives such that the longer the half-life, the smaller the increase. Protein half-lives before and after CPT addition are shown. The curved lines illustrate the predicted increase in half-lives due to reduced degradation. They are obtained by setting ki to 100%, 50%, 20% and 10% (reflecting the fold reduction in degradation) and applying Eq. 3 in 0. This pattern does not account for observed half-life increase in response to the drug.

FIG. 21 is a bar graph illustrating the null distribution generated by computing the absolute sum of differences over multiple shuffle rounds is shown, yielding a P < 10^"4. The observed differential increase in half-lives is statistically significant. The present inventors hypothesized that the longer the half-life of a protein, the larger its increase due to growth arrest. To assess the statistical significance of this assertion the following Monte-carlo shuffling test was observed. First, the deviation (absolute sum of differences) of the measured half-lives after drug from the predicted half-lives due to growth arrest was computed. Then, the changes in half-lives (Ti_/2 after drug) - (T^₂ before drug) were shuffled, and each of them was added to the pre-drug half-lives of a randomly selected protein. The null distribution generated by computing the absolute sum of differences over multiple shuffle rounds is shown, yielding a P < 10^"4.

FIG. 22 is a graph illustrating the effect of CPT of various polypeptides. A few proteins increased their removal in response to the drug CPT. The present inventors sought to identify proteins whose half-lives after the drug CPT deviated from the growth arrest model (depicted in 09A). To this end the log₂ ratio was computed between the expected and observed post-drug half-lives. The present inventors subtracted or added the .measurement error from the observed half-life of each protein so as to minimize this ratio. This correction procedure yields conservative estimations of the ratio (lower bound). The three proteins that showed the most significant deviation, CD44, DDX18 and RPS3A, increased their removal rates in response to the drug, suggesting protein specific removal regulation.

FIGs. 23A-C are graphs illustrating that the faster cells divide, the larger the expected half-life increase due to growth arrest. (A) Measured changes in half-lives (pre- and 24h post-stress) are positively correlated with changes in the corresponding protein levels. The fold-increases in protein levels compared to half-lives are smaller, possibly due to decreased production rates and the fact that protein levels have not reached steady-state. (B) The expected half -life increase is steeper the faster cells divide prior to drug arrest. Curves are for different T_cc and k₂=0 (Eq. 5). (C) The longer the protein half-life, the more sensitive it is to fluctuations in growth rate (changes in k₂, Eq. 5 - Figure 19C).

DESCRIPTION OF SPECIFIC EMBODIMENTS OF THE INVENTION

Before explaining at least one embodiment of the invention in detail, it is to be understood that the invention is not necessarily limited in its application to the details set forth in the following description or exemplified by the Examples. The invention is capable of other embodiments or of being practiced or carried out in various ways.

Analysis of post-transcription control in mammalian cells is challenging. Two methods are commonly used to determine a protein's half-life, namely radioactive pulse- chase analysis and cycloheximide chase. Pulse-chase analysis provides minimal distortion of normal cell physiology. The main disadvantages of this method are its laboriousness and necessity for radiolabeling. In contrast to pulse-chase analysis, cycloheximide chase strongly affects cellular metabolism. Importantly, both methods do not allow real-time measurements at the single cell level.

The present inventors have devised a new method for measuring the time- dependant changes of protein degradation and translation rates. This method is based on a mild bleaching of fluorescent proteins, and is simple and robust. It does not require radioactive labeling and can be scaled to measure multiple proteins at high temporal resolution in living cells.

It is important to note that this method applies regardless of how protein production rate changes in time, and does not require halting production.

Using this method, the present inventors were able to analyze the effect of a particular anti-cancer agent on the degradation rate and production rate of a variety of proteins, gleaning an abundance of information regarding target proteins. Such information may be useful in the design of novel drug therapies and the search for novel biomarkers.

Thus, according to one aspect of the present invention there is provided a method of determining a degradation rate of a fluorescently labeled polypeptide of interest in a population of cells, the method comprising:

(a) obtaining two identical populations of cells;

(d) determining fluorescence levels of the reduced fluorescently labeled polypeptide of interest in the second population of cells at the at least two different points in time;

(e) comparing fluorescence levels in the first and the second populations; and

As used herein, the phrase "degradation rate" refers to the sum of two underlying processes: intra cellular degradation (e.g. due to proteosome activity) and dilution due to cell growth that effectively reduces the protein amount by 50% every one cell generation.

The degradation rate can be mathematically calculated from the half life of a protein (ti_/2) and is equal to 1η2/ ¾_/2.

The term "cell" as used herein, refers to a biological cell, e.g. eukaryotic, such as of mammalian origin (e.g. human). The cell may be diseased (e.g. cancerous) or healthy, taken directly from a living organism or part of a cell line, immortalized or non-immortalized.

According to one embodiment, the cell is viable.

The present invention contemplates determining the degradation rate of all types of polypeptides. According to one embodiment, the polypeptide is one which is naturally expressed in the cell population. According to one embodiment, the polypeptide is not constitutively degraded at a rate faster than fluorphore maturation time, since the polypeptide would not be observable. It will be appreciated that the temporal resolution of bleach-chase is limited by the rate of the fluorescent tag folding and maturation, which ranges between a few minutes and two hours, depending on the type of fluorophore and environmental condition. Proteins that are constitutively degraded at a rate faster than fluorophore maturation time would therefore not be observable using the present assay.

The fluorescently labeled polypeptides of the present invention typically comprise a fluorescent moiety. The fluorescent moieties are typically attached covalently to the polypeptides directly (i.e. via peptide bonds), although indirect attachment via linker peptides is also contemplated.

According to an embodiment of this aspect of the invention, the fluorescent moiety is not attached to the N terminus of the polypeptide so as not to interfere with naturally occurring degradation signals.

According to one embodiment, the polypeptide maintains wild type functionality

(i.e., of non-tagged protein) and further has a similar cellular degradation pattern both prior to and following attachment of the fluorescent polypeptide. It will be appreciated however, that the method of the present invention does not require for the proteins to be functional.

Examples of suitable fluorescent moieties include, but are not limited to, phycoerythrin (PE), fluorescein isothiocyanate (FITC), Cy-chrome, rhodamine, Texas red, PE-Cy5, green fluorescent protein, the yellow fluorescent protein, the cyan fluorescent protein and the red fluorescent protein as well as their enhanced derivatives.

Table 1 below provides examples of sequences of identifiable moieties.

Table 1

Identifiable Moiety Amino Acid sequence Nucleic Acid sequence

(Genebank Accession No.) (Genebank Accession No.)

Green Fluorescent protein AAL33912 AF435427

orange fluorescent protein AAL33917 AF435432

Fluorescein isothiocyanate AAF22695 AF098239 For additional guidance regarding fluorophore selection, methods of linking fluorophores to various types of molecules see Richard P. Haugland, "Molecular Probes: Handbook of Fluorescent Probes and Research Chemicals 1992-1994", 5th ed., Molecular Probes, Inc. (1994); U.S. Pat. No. 6,037,137 to Oncoimmunin Inc.; Hermanson, "Bioconjugate Techniques", Academic Press New York, N.Y. (1995); Kay M. et al, 1995. Biochemistry 34:293; Stubbs et al, 1996. Biochemistry 35:937; Gakamsky D. et al, "Evaluating Receptor Stoichiometry by Fluorescence Resonance Energy Transfer," in "Receptors: A Practical Approach," 2nd ed., Stanford C. and Horton R. (eds.), Oxford University Press, UK. (2001); U.S. Pat. No. 6,350,466 to Targesome, Inc.].

Methods of generating fluorescently labeled polypeptides in cells are known in the art. For example, polypeptides may be expressed as fusion proteins in cells using recombinant DNA technology. The DNA sequence encoding the fusion proteins is inserted into nucleic acid constructs and cells are trasfected using methods commonly known in the art as described further herein below.

Examples of mammalian expression vectors include, but are not limited to, pcDNA3, pcDNA3.1 (+/-), pGL3, pZeoSV2(+/-), pSecTag2, pDisplay, pEF/myc/cyto, pCMV/myc/cyto, pCR3.1, pSinRep5, DH26S, DHBB, pNMTl, pNMT41, and pNMT81, which are available from Invitrogen, pCI which is available from Promega, pMbac, pPbac, pBK-RSV and pBK-CMV, which are available from Strategene, pTRES which is available from Clontech, and their derivatives.

According to one embodiment the fluorescently labeled polypeptide is transcribed from its native chromosomal location (i.e. it is endogenous to the cell).

Various methods are contemplated for inserting a fluorescent moiety into the genome of the host cell in order to allow transcription of the polypeptide from its native chromosomal location. Such methods include for example homologous recombination, site-specific recombination non-homologous recombination.

As used herein, the phrase "homologous recombination" refers to the process in which nucleic acid molecules with similar nucleotide sequences associate and exchange nucleotide strands. A nucleotide sequence of a first nucleic acid molecule that is effective for engaging in homologous recombination at a predefined position of a second nucleic acid molecule will therefore have a nucleotide sequence that facilitates the exchange of nucleotide strands between the first nucleic acid molecule and a defined position of the second nucleic acid molecule. Thus, the first nucleic acid will generally have a nucleotide sequence that is sufficiently complementary to a portion of the second nucleic

As used herein, the phrase "site-specific recombinase" refers to a type of recombinase that typically has at least the following four activities (or combinations thereof): (1) recognition of specific nucleic acid sequences; (2) cleavage of the sequence or sequences; (3) topoisomerase activity involved in strand exchange; and (4) ligase activity to reseal the cleaved strands of nucleic acid (see Sauer, B., Current Opinions in Biotechnology 5:521-527 (1994)). Conservative site-specific recombination is distinguished from homologous recombination and transposition by a high degree of sequence specificity for both partners. The strand exchange mechanism involves the cleavage and rejoining of specific nucleic acid sequences in the absence of DNA synthesis (Landy, A. (1989) Ann. Rev. Biochem. 58:913-949).

Nucleic acid constructs (also referred to herein as "expression vectors") capable of insertion in a directed manner typically comprise one or more functionally compatible recognition site for a site-specific recombination enzyme.

As used herein, the phrase "functionally compatible recognition sites for a site- specific recombination enzyme" refers to specific nucleic acid sequences which are recognized by a site-specific recombination enzyme to allow site-specific DNA recombination {i.e., a crossover event between homologous sequences). An example of a site-specific recombination enzyme is the Cre recombinase (e.g., GenBank Accession No. YP_006472), which is capable of performing DNA recombination between two loxP sites. Cre recombinase can be obtained from various suppliers such as the New England BioLabs, Inc, Beverly, MA, or it can be expressed from a nucleic acid construct in which the Cre coding sequence is under the transcriptional control of an inducible promoter (e.g., the galactose-inducible promoter) as in plasmid pSH47.

The phrase, "non-homologous recombination" as used herein refers to the joining (exchange or redistribution) of genetic material through a mechanism that does not involve homologous recombination (e.g., recombination directed by sequence homology) and that does not involve site-specific recombination (e.g., recombination directed by site-specific recombination signals and a corresponding site-specific recombinase). Examples of non-homologous recombination include integration of exogenous DNA into chromosomes at non-homologous sites, chromosomal translocations and deletions, DNA end joining, double strand break repair, bridge-break- fusion, concatemerization of transfected polynucleotides, retroviral insertion, and transposition.

Retroviral vectors integrate into eukaryotic genomes by a distinct mechanism of non-homologous recombination that is catalyzed by the action of the virally encoded integrase enzyme, and the mechanism of viral integration, replication and infection has been well described [see for example Retroviruses. Coffin, J M.; Hughes, S H.; Varmus, H E. Plainview (NY): Cold Spring Harbor Laboratory Press; cl997; Use of wildtype retroviruses as mutagens]. The mutagenic ability of retroviruses and retroviral vectors and their ability to enable the rapid identification of mutated genes through the linkage of retroviral tag sequences within the transcripts of mutagenized genes are well known in the art (Friedrich G, Soriano P. Methods Enzymol. 1993;225:681-701; 3: Gossler A, et al., Science. Apr. 28, 1989;244(4903):463-5; Friedrich G, Soriano P. Genes Dev. September 1991;5(9):1513-23; 5: von Melchner H, et al Genes Dev. June 1992;6(6):919-27].

Retroviral constructs of the present invention may contain retroviral LTRs, packaging signals, and any other sequences that facilitate creation of infectious retroviral vectors. Retroviral LTRs and packaging signals allow the reporter polypeptides of the invention to be packaged into infectious particles and delivered to the cell by viral infection. Methods for making recombinant retroviral vectors are well known in the art (see for example, Brenner et al., PNAS 86:5517-5512 (1989); Xiong et al., Developmental Dynamics 212:181-197 (1998) and references therein; each incorporated herein by reference). In preferred embodiments, the retroviral vectors used in the invention comprise splice acceptor (SA) and splice donor (SD) sequences flanking the sequence encoding the reporter polypeptide. Typically, the constructs of the present invention do not comprise a promoter, a start codon or a poly A signal. In this way, if the virus inserts into an actively transcribed gene, the reporter sequence is retained as a new exon after splicing of the mRNA. Owing to the large size of the first intron and viral preference for integration sites near the start of genes, the first intron is the most common point of insertion. The tagged mRNA translates to an internally labeled protein, with the reporter polypeptide usually near the N terminus.

Retroviral LTRs and packaging signals can be selected according to the intended host cell to be infected. Examples of retroviral sequences useful in the present invention include those derived from Murine Moloney Leukemia Virus (MMLV), Avian Leukemia Virus (ALV), Avian Sarcoma Leukosis Virus (ASLV), Feline Leukemia Virus (FLV), and Human Immunodeficiency Virus (HIV). Other viruses known in the art are also useful in the present invention and therefore will be familiar to the ordinarily skilled artisan.

Like retroviruses, transposons and transposon vectors can also be used to integrate sequences in a non-directed fashion into the chromosome of the cell. Also like retroviruses, transposons integrate by enzymatically catalyzed non-homologous recombination in which transposase enzymes catalyze the genomic integration and transposition of transposon DNA.

Numerous transposons have been characterized that function in mammals. In particular, the TCI/mariner derivative transposon, Sleeping Beauty, has been demonstrated to integrate efficiently in mammals.

When the fluorescent polypeptide is inserted into the genome in a non-directed fashion, the fluorescent moiety may be identified, such as by 3' RACE, using a nested PCR reaction that amplifies the section between the reporter polypeptide and the polyA tail of the mRNA of the host gene. The PCR product may be sequenced directly and aligned to the genome, thereby identifying the polypeptide.

Exemplary oligonucleotide primers that may be used for 3 'RACE and sequencing are listed in Table 2 herein below.

Table 2

Primer name Use Sequence Alignment in YFP or mCherry

AP first-strand First-strand cDNA GGCCACGCGTCGACTAGTAC(T)17

synthesis (SEQ ID NO: 23)

AP 92 RACE first and GGCCACGCGTCGACTAGTAC

nested reaction 3' (SEQ ID NO:24)

primer

YFP 90 RACE first GCAGAAGAACGGCATCAAGG Bases 471^90 reaction 5' primer (SEQ ID NO: 25)

for YFP-tagged genes

YFP 85 RACE-nested CGCGATCACATGGTCCTGCTG Bases 646-666 reaction 5' primer (SEQ ID NO: 26)

for YFP-tagged

genes

Cherry 45 RACE first GTGGTGACCGTGACCCAGGA Bases 322-341 reaction 5' primer (SEQ ID NO: 27)

for mCherry-tagged

genes

Cherry 46 RACE-nested GCGGATGTACCCCGAGGACG Bases 456-475 reaction 5<2 primer (SEQ ID NO: 28)

for mCherry-tagged

genes

Cherry 56 Sequencing of GACTACACCATCGTGGAACA Bases 586-605 mCherry RACE (SEQ ID NO: 29)

product

YFP 906 Sequencing of YFP GGATCACTCTCGGCATGGAC Bases 686-705

RACE product (SEQ ID NO: 30)

The constructs of the present invention can be introduced into a cell and integrated into DNA by any method known in the art. In one embodiment, they are introduced by transfection. Methods of transfection include, but are not limited to, electroporation, particle bombardment, calcium phosphate precipitation, lipid-mediated transfection (e.g., using cationic lipids), micro-injection, DEAE-mediated transfection, polybrene mediated transfection, naked DNA uptake, and receptor mediated endocytosis.

Typically the introduction of the constructs of the present invention is effected whilst the cells are being cultured in a medium which supports well-being and propagation. The medium is typically selected according to the cell being transfected/infected.

According to one embodiment, the constructs of the present invention are introduced into the cell by viral transduction or infection. Suitable viral vectors useful in the present invention include, but are not limited to, adeno-associated virus, adenovirus vectors, alpha-herpesvirus vectors, pseudorabies virus vectors, herpes simplex virus vectors and retroviral vectors (including lentiviral vectors).

In order to carry out the method of the present invention, two identical populations of cells are required each expressing the fluorescently labeled polypeptide to the same extent.

The phrase "identical cell populations" refers to homogeneous cell populations (i.e. of the same cell type) and being under the same conditions (e.g. culturing conditions), wherein the proteins in each of the populations are identically labeled. Thus for example one population of cells may be obtained which expresses the fluorescently labeled polypeptide which is then divided into at least two sub-populations.

As mentioned, the method of the present invention is effected by determining the change in fluorescence levels of the fluorescently labeled polypeptide over time in *both the cell populations. The measurements may be effected simultaneously, or one following the other in any order.

Methods of analyzing fluorescence levels are known in the art. Preferably, the analysis does not affect the viability or function of the cell. Fluorescence levels may be measured using a fluorescent confocal microscope or using flow cytometry at particular time points. Alternatively, fluorescence levels may be measured in real-time using long period time-lapse microscopy. Time-lapse movies may be obtained as described by Sigal et al. (Sigal, Milo et al. 2006, supra) with for example an automated, incubated (including humidity and C0₂ control) inverted fluorescence microscope (e.g. Leica DMIRE2) and a CCD camera (e.g. ORCA ER- Hamamatsu Photonics).

According to one embodiment the measurements are taken over a period of an hour, over a period of 6 hours, over a period of 12 hours, over a period of 18 hours, over a period of 24 hours or over a period of 48 hours. The number of measurements is not limited by typically are in the range of 2-20.

Prior to measuring fluorescence in the second cell population, the cells are treated in order to reduce fluorescence. According to one embodiment the reducing transforms a fraction of the fluorescent proteins into non-fluorescent. This action is referred to herein as bleaching. .

According to one embodiment, the bleaching is such that there is a 10 - 100 % reduction, more preferably a 10-90 % reduction, more preferably a 10-80 % reduction, more preferably a 10-70 % reduction, more preferably a 10-60 % reduction, more preferably a 10-50 % reduction, in the level of fluorescence.

Preferably, the bleaching is irreversible, or at least irreversible over the time over which measurements are recorded.

According to another embodiment, the bleaching does not alter cell -viability, motility, mitosis rate or morphology. An exemplary method of bleaching includes shining pulses of light on the cells. The pulses of light can last from 15 seconds to 30 minutes (e.g. 1-8 minutes).

Following fluorescence measurements, the change in fluorescence levels in the bleached cell is compared to the change in fluoresecence levels in the non-bleached cells.

The present inventors have shown that the degradation rate is equal to the ratio of (the natural log (In) of the difference in fluorescence of the polypeptide in the non- bleached cells and fluorescence of the polypeptide in the bleach cells at a first point in time) : (natural log (In) of the difference in fluorescence of the polypeptide in the non- bleached cells and fluorescence of the polypeptide in the bleached cells at a second point in time) divided by the difference in time between the two points in time.

For a more accurate determination of the degradation rate fluorescence levels may be measured over a number of time points. A graph can then be plotted, with the y axis indicating the In (P(t) -Pv(t)) and the x axis graph indicating time, wherein P(t) is the amount of fluorescence of the polypeptide at point t in time in the non-bleached cells and (Pv(t) is the amount of fluorescence of polypeptide at point t in time in the bleached population of cells. The slope of the graph indicates the degradation rate.

It will be appreciated that in the case where the degradation rate is not constant it may be retrieved by using the following formula.

t

In (P(t) - P_v(t)) = In (P₀-P_0v) - J a(t') df where P and P0 are two measurements of the polypeptide without the bleaching at two different time points. Pv and PvO are two measurements of the polypeptide with the bleaching at two different time points.

is the integral over the two different time points.

It will further be appreciated that once the degradation rate of a polypeptide is calculated, it is also possible to calculate the production rate. Measurement of production rate is only possible when the fluorescently labeled polypeptide is transcribed from its native chromosomal location, as described herein above. Thus according to another aspect of the present invention, there is provided a method of determining a production rate of the polypeptide. The production rate may be determined according to the following equation: Eq. (2) β(ί) = dP(t)/dt + a(t) . P(t) where beta is the production rate at a given time point (t). dP(t)/dt is equal to the change in protein in time, alpha is the degradation rate at a given point in time and P is the protein level at a given point in time.

Various methods of measuring protein levels are known in the art including quantitative immunoblots, Bradford assay etc.

According to one embodiment, protein levels are measured by causing the cells to express an additional fluorescently labeled polypeptide in order to eliminate variations in lamp intensities across experiments. According to an embodiment of this aspect of the present invention the additional fluorescently labeled polypeptide has a higher nuclear: cytoplasm expression ratio and further is distinguishable from the fluorescently labeled polypeptide of interest. Thus, fluorescent moieties for example may be selected such that each emits light of a distinguishable wavelength and therefore color when excited by light.

Once protein levels P(t) are measured at different time points, one can retrieve dP(t)/dt by subtracting the protein levels in consecutive times points

(i.e. dPi/dt = Pi+1 - Pi).

The method of this aspect of the present invention may be adapted in order to test the effect of a particular agent on the degradation rate of the cells. Thus, according to another aspect of the invention there is provided a method of analyzing the effect of an agent on a degradation rate of a polypeptide of interest in a population of cells, the method comprising:

(a) contacting a population of cells with the agent;

(b) determining the degradation rate of the fluorescently labeled polypeptide of interest in the population of cells according to the method of claim 1; and (c) comparing the degradation rate obtained in step (b) with a degradation rate of the polypeptide in an absence of the agent, thereby analyzing the effect of the agent on the degradation rate of the polypeptide of interest.

According to this aspect of the present invention the degradation rate of the polypeptide in the absence of the agent may be known in the literature or ascertained using any of the methods known in the art.

It will be appreciated that the method of the present invention may also be adapted to compare the degradation rate of a polypeptide in diseased and healthy cells.

It will be further appreciated that the method disclosed herein can be used to study protein degradation in different cellular compartments (e.g. nucleus versus cytoplasm) because bleach-chase allows a microscopy-based assay of protein removal. Additionally, bleach-chase facilitates measurements of degradation rates as a function of cell-cycle stage, without the need for chemical or physical synchronization. This can be achieved by applying bleach-chase to a population of unsynchronized cells, followed by in-silico synchronization.

The conversion from rate of change in fluorescence of a polypeptide under bleached and non-bleached conditions to degradation rate and/or production rate of the polypeptide may be effected manually or using a computer system comprising a processing unit executing a software application configured for converting a difference in rate of change of fluorescence between a bleached and non-bleached population of cells to degradation rate and/or production rate.

As used herein, the phrase "processing unit" refers to a data processor, e.g., a computer.

The software application can be embodied in a tangible medium such as, but not limited to, a floppy disk, a CD-ROM, a hard drive of a computer, and a memory medium (e.g., RAM, ROM, EEPROM, flash memory, etc.). The software can be run by loading the software into the execution memory of the processing unit, configuring the processing unit to act in accordance with the instructions of the software. All these operations are well-known to those skilled in the art of computer systems.

According to an embodiment of this aspect of the present invention, the system may further comprise an imaging system. Such an imaging system typically comprises an illuminating device for the purposes of measuring fluorescence. The illuminating device may also comprise a white light source which may be required for delineating the borders of the sample being analyzed. Alternatively or additionally, the illuminating device may comprise light sources of different wavelengths such that it is capable of detecting more than one type of fluorescent moiety.

The imaging system of the present invention further comprises an image capture apparatus. The image capture apparatus typically includes means to acquire the image (e.g. a CCD) by translating the light into electronic impulses and transmits the image to a display device. The image capture apparatus may also include means for magnifying the image (e.g., a microscope). The CCD may be any suitable photosensitive device, including but not limited to a photomultiplier tube (PMT), a phototransistor, or a photodiode.

Examples of picture elements include, but are not limited to a pixel or a group of pixels.

According to a preferred embodiment of this aspect of the present invention, the display device of the system displays the degradation rate and/or the production rate. The display may be in numerical form and/or graphical form.

According to an additional embodiment of this aspect of the present invention, the system further comprises a time recording device.

According to a preferred embodiment of this aspect of the present invention, the image capture apparatus is calibrated for capturing an image of the fluorescently labeled polypeptide. The calibrations typically take into account the cell population (e.g. number of cells, size of cells) and the particular fluorescent moiety used.

As used herein the term "about" refers to ± 10 %

The terms "comprises", "comprising", "includes", "including", "having" and their conjugates mean "including but not limited to".

The term "consisting of means "including and limited to".

The term "consisting essentially of" means that the composition, method or structure may include additional ingredients, steps and/or parts, but only if the additional ingredients, steps and/or parts do not materially alter the basic and novel characteristics of the claimed composition, method or structure. As used herein, the singular form "a", "an" and "the" include plural references unless the context clearly dictates otherwise. For example, the term "a compound" or "at least one compound" may include a plurality of compounds, including mixtures thereof.

Throughout this application, various embodiments of this invention may be presented in a range format. It should be understood that the description in range format is merely for convenience and brevity and should not be construed as an inflexible limitation on the scope of the invention. Accordingly, the description of a range should be considered to have specifically disclosed all the possible subranges as well as individual numerical values within that range. For example, description of a range such as from 1 to 6 should be considered to have specifically disclosed subranges such as from 1 to 3, from 1 to 4, from 1 to 5, from 2 to 4, from 2 to 6, from 3 to 6 etc., as well as individual numbers within that range, for example, 1, 2, 3, 4, 5, and 6. This applies regardless of the breadth of the range.

Whenever a numerical range is indicated herein, it is meant to include any cited numeral (fractional or integral) within the indicated range. The phrases "ranging/ranges between" a first indicate number and a second indicate number and "ranging/ranges from" a first indicate number "to" a second indicate number are used herein interchangeably and are meant to include the first and second indicated numbers and all the fractional and integral numerals therebetween.

As used herein the term "method" refers to manners, means, techniques and procedures for accomplishing a given task including, but not limited to, those manners, means, techniques and procedures either known to, or readily developed from known manners, means, techniques and procedures by practitioners of the chemical, pharmacological, biological, biochemical and medical arts.

It is appreciated that certain features of the invention, which are, for clarity, described in the context of separate embodiments, may also be provided in combination in a single embodiment. Conversely, various features of the invention, which are, for brevity, described in the context of a single embodiment, may also be provided separately or in any suitable subcombination or as suitable in any other described embodiment of the invention. Certain features described in the context of various embodiments are not to be considered essential features of those embodiments, unless the embodiment is inoperative without those elements.

Various embodiments and aspects of the present invention as delineated hereinabove and as claimed in the claims section below find both calculated and experimental support in the following examples.

EXAMPLES

Reference is now made to the following examples, which together with the above descriptions illustrate some embodiments of the invention in a non limiting fashion.

Generally, the nomenclature used herein and the laboratory procedures utilized in the present invention include molecular, biochemical, microbiological and recombinant DNA techniques. Such techniques are thoroughly explained in the literature. See, for example, "Molecular Cloning: A laboratory Manual" Sambrook et al., (1989); "Current Protocols in Molecular Biology" Volumes I-III Ausubel, R. M., ed. (1994); Ausubel et al., "Current Protocols in Molecular Biology", John Wiley and Sons, Baltimore, Maryland (1989); Perbal, "A Practical Guide to Molecular Cloning", John Wiley & Sons, New York (1988); Watson et al., "Recombinant DNA", Scientific American Books, New York; Birren et al. (eds) "Genome Analysis: A Laboratory Manual Series", Vols. 1-4, Cold Spring Harbor Laboratory Press, New York (1998); methodologies as set forth in U.S. Pat. Nos. 4,666,828; 4,683,202; 4,801,531; 5,192,659 and 5,272,057; "Cell Biology: A Laboratory Handbook", Volumes I-III Cellis, J. E., ed. (1994); "Culture of Animal Cells - A Manual of Basic Technique" by Freshney, Wiley- Liss, N. Y. (1994), Third Edition; "Current Protocols in Immunology" Volumes I-III Coligan J. E., ed. (1994); Stites et al. (eds), "Basic and Clinical Immunology" (8th Edition), Appleton & Lange, Norwalk, CT (1994); Mishell and Shiigi (eds), "Selected Methods in Cellular Immunology", W. H. Freeman and Co., New York (1980); available immunoassays are extensively described in the patent and scientific literature, see, for example, U.S. Pat. Nos. 3,791,932; 3,839,153; 3,850,752; 3,850,578; 3,853,987; 3,867,517; 3,879,262; 3,901,654; 3,935,074; 3,984,533; 3,996,345; 4,034,074; 4,098,876; 4,879,219; 5,011,771 and 5,281,521; "Oligonucleotide Synthesis" Gait, M. J., ed. (1984); "Nucleic Acid Hybridization" Hames, B. D., and Higgins S. J., eds. (1985); "Transcription and Translation" Hames, B. D., and Higgins S. J., eds. (1984); "Animal Cell Culture" Freshney, R. L, ed. (1986); "Immobilized Cells and Enzymes" IRL Press, (1986); "A Practical Guide to Molecular Cloning" Perbal, B., (1984) and "Methods in Enzymology" Vol. 1-317, Academic Press; "PCR Protocols: A Guide To Methods And Applications", Academic Press, San Diego, CA (1990); Marshak et al., "Strategies for Protein Purification and Characterization - A Laboratory Course Manual" CSHL Press (1996); all of which are incorporated by reference as if fully set forth herein. Other general references are provided throughout this document. The procedures therein are believed to be well known in the art and are provided for the convenience of the reader. All the information contained therein is incorporated herein by reference.

GENERAL MATERIALS AND METHODS

Fluorescently-tagged protein clones: Clones used in this study were taken from the LARC library, in which proteins were fluorescently-tagged under their endogenous regulation by CD-tagging as previously described [14, 17-18, 30-31]. Briefly, Clones are based on H1299 non small cell lung carcinoma where each clone contains two tags: The first, common to all clones, is a red fluorescent protein mCherry that creates a pattern, which is strong in the nucleus and weaker in the cytoplasm and is used for image analysis purposes. The second is a yellow tag (eYFP or Venus) of the protein of interest. The CD-tagging scheme used to generate the clones tends to preserve protein functionality and localization [15, 18, 27-29]. Note, however that the purpose of this study does not require the proteins to be functional but merely reliable reporters of the endogenous protein dynamics. Additional information regarding the LARC library is published elsewhere [18].

Tissue culture media: Cells were grown in RPMI 1640 supplied with (+) L-

Glutamine (GIBCO, cat. No. 21875) medium supplemented with 10 % Fetal Calf Serum (certified fetal bovine serum, membrane filtered, Biological Industries, 04-001-lA) and 0.05 % Penicillin-Streptomycin antibiotics (Biological Industries Cat. No. 03-031-1B). Cells were grown in incubators at 37 °C and 8 % C0₂.

CPT drug addition: Camptothecin (CPT, C9911 Sigma) was dissolved in

DMSO giving a stock solution of 10 mM, In each experiment, the drug was diluted to the desired concentration in transparent growth medium (RPMI 1640, 0.05 % Penicillin- Streptomycin antibiotics, 10 % FCS, with L-Glutamine, lacking riboflavin and phenol red, Bet Haemek, Biological Industries Cat. No.06-1100-26-lA). Normal transparent growth medium (2 ml, without the drug) was replaced by the diluted drug (2 ml medium with 10 μΜ CPT diluted 1:1000 from stock) under the microscope after at least one round without the drug.

Time-lapse microscopy: Time-lapse movies were obtained at 20x magnification. Three automated microscopes were used, based on inverted fluorescence microscopes from Leica (DMIRE2 and DMI6000B). All microscopes included live cell environmental incubators maintaining 37 °C (37-2 digital and Heating unit, PeCon,Germany, Leica #15531719) humidity and 8 % C0₂ (PeCon, GmbH, Germany #0506.000-230, Leica #11521733) and automated stage movement control (Corvus, ITK, GmbH, Germany); stage was surrounded by a custom enclosure to maintain constant temperature, C0₂ concentration, and humidity. Transmitted and fluorescent light paths were controlled by electronic shutters (Uniblitz, model VMM-D1, Rochester, NY); Fluorescent light sources were Mercury short arc lamp HXP and Mercury HBO100 (OSRAM, Germany). Cooled 12 and 14 bit CCD cameras were used Qlmaging, (RETIGA-S&V, Fast 1394, RET-SRV-F-M-12-C, Canada); CoolSNAP, (Roper Scientific HQ, photometries); ORCA-ER (C4742-95-12ERG, and Hamamatsu photonics K.K, Japan). The filters used were from Chroma Technology Corp: single channel filters were YFP: (500/20 nm excitation, 515 nm dichroic splitter, and 535/30 nm emission, Chroma #41028) and mCherry Red: (575/50 nm excitation, 610 nm dichroic splitter, and 640/50nm emission, Chroma #41043); The hardware was controlled by ImagePro5 Plus software (Media Cybernetics) with integrated time-lapse acquisition, stage movement, and software based auto-focus.

Cells were grown and visualized in 12-well optical glass-bottom plates (MatTek cultureware, Microwell plates-uncoated, part No. P12G-0-14-F, Lot No. TK0289) coated with 10μΜ fibronectin 0.1% (solution from bovine plasma, Sigma, Cat. No. F1141) diluted 1:100 in Dulbecco's Phosphate Buffered Saline, PBS (Sigma, Cat. No. D8537). In order to reduce background fluorescence, cells were gown in RPMI 1640 lacking riboflavin and phenol red (Beit Haemek, Biological Industries Cat. No.06-1100- 26-1 A), supplemented with L-Glutamine, 10% Fetal Calf Serum (certified fetal bovine serum, membrane filtered, Biological Industries, 04-001-lA) and 0.05% Penicillin- Streptomycin antibiotics (Biological Industries Cat. No. 03-031 -IB). For each well, time-lapse movies were obtained at four fields of view. Each movie was taken at a time resolution of 20 minutes and was filmed for at least two days (over 140 time points). Each time point included transmitted light image (phase contrast), and two fluorescent channels (red and yellow). No significant bleaching was observed (on average less than 3 % over the duration of the experiment)

Image analysis of time lapse movies: Custom computer vision software was built to automatically extract quantitative measurements of protein dynamics from time- laps movies. The main modules in the software include background normalization, cell segmentation, cell tracking and automated identification of various cellular phenotypes (e.g. mitosis and cell death). Cell and nuclei segmentation was based on the patterns generated by the red-tagging common to all clones. An LDA classification approach [32] is first used to identify cell nucleus according to intensity and texture followed by seeded watershed segmentation in order to identify cell boundaries [33]. The next step, cell tracking, links each cell in a given frame to the appropriate cell in the preceding and following frames. To this end the following scheme was applied: each triplet of cells (i.e. a cell in frame i - 1 and / + 1) generates a trajectory whose cost depends on its velocity consistency (large fluctuations in velocity increase the cost) angle smoothness (abrupt changes in trajectory angle increase the cost) as well as the similarity between the size and appearance of the connected cells. The problem of tracking is then reduced to finding a non-overlapping assignment of trajectories which minimizes the global trajectory cost [34]. Lastly, the software identifies different cellular phenotypes such as apoptosis and mitosis using a machine learning approach [18].

Protein Calibration protocol: A method of protein calibration was used to compare levels across different proteins. The cherry fluorescence marker was use, common to all the clones used in this study, as a common reference in order to eliminate variations in lamp intensities across experiments. The levels of both yellow and red fluorescence measurements are described by the following two equations:

Eq. (7) F_y = P_y . T_y . I

Eq. (8) F_r = P_r . T_r . I

SUBSTITUTE SHEET RULE where P_x is the number of tagged protein copies, T_x is the exposure time, I is the intensity of the lamp and sub-index y and r denote YFP and cherry respectively. It follows that the lamp intensity is:

Eq. (9) I = F_r/ P_r . T_r

By applying Eq. 9 on Eq. 7 and 8 and omitting P_r (since it is constant in all clones) one finds:

Eq. (10) P_y = F_y. TV F_r. T_y

Eq. 10 is used to retrieve the levels of the YFP-tagged protein. Briefly the reproducibility of this approach was tested on 13 proteins spanning a wide range of expression levels (over 500 fold ratio between lowest and highest levels) and found it better than 30% (std/mean). This approach was further tested by comparing protein fluorescence levels to quantitative immunoblots with protein standards (no. of replicates > 3). Measurements obtained by the two techniques are in good agreement (r > 0.8, P < 0.001). '

Microarray experiments: Cells were plated in 10 cm plates (1.6xl0⁴ cells per plate) with at least three replicates for each time point. CPT (10 μΜ) was added 48 hours after plating. At each time point (0, 0.5, 1.5, 3, 6, 9, 12, 15, 18, 23, 28, 33, 38, 43 and 48 hours after drug addition) cells were trypsinized and counted followed by lyses of 4xl0⁴ and total RNA was purification using RNeasy kit (QIAGEN), with elution in 30 μΐ of RNase-free water. A total of 10 μΐ of 200 ng^l (66.67 ng/μΐ) RNA were transferred to microarray chip hybridization. Amplified cRNA was prepared from 200 ng total RNA using the WT cDNA Synthesis and WT cDNA Amplification Kits (900672, Affymetrix). Biotinylated single-stranded cDNA was generated from the amplified cRNA and then fragmented and labeled with the WT Terminal Labeling Kit (Affymetrix), following manufacturer protocol. Samples were hybridized to Human Gene 1.0 ST Arrays (Affymetrix) and scanned at the Weizmann Institute Microarray Core Facility using the Affymetrix GeneChip Scanner 3000 7G. Raw data is given in the SOM. Partek Genomic Suite software was then used to extract raw data, perform mean probe summarization, RMA and quintile normalization and GC content correction [35].

SUBSTITUTE SHEET (RULE 26} Quantitative real-time PCR: To validate the microarray measurements we selected 10 genes that exhibited different drug response profiles and measured their mRNA using qPCR (TOPI, PSMB4, AIP, CKS, JAGN1, DDX5, ENOl, NCBP2, MYH9 and GAPDH). To this end, cells were plated at 1.6xl0⁴ cells in a 10cm plate with five replicates at each time point. 48 hours after plating, 10μΜ of Camptothecin(CPT) was added. Cells were harvested at different time points after CPT addition (0, 2, 9, 23, 33 and 48 hours). Then, 5xl0⁴ cells were lyzed and total RNA was purified using the RNeasy kit (QIAGEN), followed by quantification using the NanoDrop machine (ND-1000 Spectrophotometer). RNA samples were then reverse transcribed by Omniscript RT kit (Qiagen) and random hexamers. Real-Time PCR reactions were performed in a 20 μΐ mixture containing 1/10 volume of cDNA (2 μΐ), and 18 μΐ of Real-time PCR reaction mix containing 10 μΐ SYBR Green buffer (PE Applied Biosystems, Foster City, CA, USA), 0.2 μΜ of each primers (forward and reverse of each gene), and 7.6μ1 DDW. QRT-PCR reactions were performed on a qRTPCR Strategene Mx 3000P machine. Primer sequences using for the qPCR are provided in Table 3, herein below.

Table 3

Gene symbol Forward Primer (5' to 3 ') Reverse Primer (5' to 3 ')

TOPI TCCGGAACCAGTATCGAGAAGA CCTCCTTTTCATTGCCTGCTC

(SEQ ID NO: 1) (SEQ ID NO: 2)

PSMB4 ACGCGGACCCAGAACCCCAT GCGTAGTCGCCAGAGGCACC

(SEQ ID NO: 3) (SEQ ID NO: 4)

AIP GCACTGCTGCGGTGTTGCAC TGCCAGGGCTCTCCACCTTCA

(SEQ ID NO: 5) (SEQ ID NO: 6)

GAPDH AGGCTGGGGCTCATTTGCAGG TGACCTTGGCCAGGGGTGCT

(SEQ ID NO: 7) (SEQ ID NO: 8)

CKS2 GTTGCCTGGGCTGGACGTGG AACATGCCGGTACTCGTAGTGTTCG

(SEQ ID NO: 9) (SEQ ID NO: 10)

JAGN1 TCTGGGCAGGCACAATGGCG AGCCTCAGGTGTCCCACGCT

(SEQ ID NO: 11) (SEQ ID NO: 12)

PPIA GCCAAGACTGAGTGGTTGGATGGC TGGGGTGGAGGGGTGCTCTC

(SEQ ID NO: 13) (SEQ ID NO: 14) DDX5 GGAGTGCGACTTGGCCAAAAGA GGCAGGCCACCCATCTCTCCT (SEQ ID NO: 15) (SEQ ID NO: 16)

ENOl TGGTGTCTATCGAAGATCCCTT CACTGTGAGATCATCCCCCAC

(SEQ ID NO: 17) (SEQ ID NO: 18)

NCBP2 TGCCTTGACATGAGGACCAGCA GGCATCTGCTGCAGGAACCTT

(SEQ ID NO: 19) (SEQ ID NO: 20)

MYH9 AAGCAGGCGTGCGTGCTCAT CTGCAGCAGGCCTGGAACCC

(SEQ ID NO: 21) (SEQ ID NO: 22)

GO Enrichment analysis: GO enrichment analysis was applied in order to identify groups of genes exhibiting similar dynamic profiles in response to the drug that share a common function, process or localization [36]. To this end Gorilla was applied, a software for automatic identification and visualization of enriched GO terms in ranked lists of genes [37]. For each time point, genes were ranked according to their level of expression (over-expressed at the top and under-expressed at the bottom of list). This was given as input to GOrilla, which identifies sets of genes sharing a common function, process or localization that are enriched at the top of the list. The determination of the exact cutoff is done automatically, using an optimization scheme called mHG which produces an enrichment P -value [37-38]. Enrichment was considered significant if a term showed a P < 10^"4 after multiple hypothesis correction due to the multiple time points and GO terms. The same procedure was then repeated on the inverse lists identifying sets of genes sharing similar functions that were down regulated at some time point in response to the drug. A summary of enriched GO terms as a function of time after drug addition is shown in Figures 1A-K.

Time shift correlations: To allow for effects of delay between mRNA and protein, the temporal protein profiles were shifted with respect to the mRNA profiles. Using a sliding window ranging from -10 to +10 hours with 20 minutes resolution the Pearson correlation was recorded at each time point. Alignment of mRNA and protein profiles was determined according to the time point at which maximal correlation was attained.

Pulse chase protocol for measuring protein half-lives: Approximately 2xl0⁶ cells were plated in 10 cm dishes and cultured for 24 hours as described above. To perform the pulse-chase, cell monolayers were rinsed twice with warm sterile PBS and starved of methionine and cysteine by incubation for 1 hour at 37 °C in 5 ml of Pulse- chase Medium: methionine/cysteine-free RPMI 1640 (Sigma, R-7513) supplemented with 10 % dialyzed FBS (Biological Industries Beit Haemek) and 2 mM L-glycine. Following starvation, the cells were labeled for 1 hour at 37 °C with 4 ml of Pulse-chase medium containing 4 mCi (lmCi/mL) of [35S]methionine/cysteine (EasyTag™Expre³⁵S³⁵S Protein Labeling Mix, #PENEG772, PerkinElmer). The radioactive medium was then removed, the cells rinsed three times with PBS, re-fed with Pulse-chase medium supplemented with 2.5 mM L-Methionine and 2.5 mM L- Cysteine Hydrochloride Hydrate (Biological Industries Beit Haemek), and incubated for the indicated times. The cells were then collected and lysed with modified RIPA buffer (50 mM, Tris-HCl , pH 7.4; 150 mM NaCl; 1 mM EDTA; 1 % NP-40; 0.25 % Na- deoxycholate; 1 mM PMSF and 1 g/ml of each aprotinin, leupeptin and pepstatin). The protein concentrations were determined by BCA protein assay kit (Thermo scientific). The proteins of interest were recovered from 200 of cell lysates by immunoprecipitation using anti-GFP or protein-specific antibodies listed below. The immunoprecipitates were subjected to SDS-PAGE, and protein bands were detected by autoradiography. Parallel gels were blotted onto nitrocellulose membrane and probed with protein-specific antibodies to ensure correct identification of the protein bands of interest.

It is assumed that radioactive protein, P(t), decays in time at a rate that is exponential in the degradation rate, a:

Eq. (ll) P(t) = P₀e ^at,

where P₀ is protein levels at t = 0 and hence one can retrieve a by applying a linear regression to the following equation:

Eq. (12) ln(P(t)) = ln(P0) - a(t) . t.

Applying the translation inhibitor CHX for measuring protein half-lives:

Cyclohexamide (CHX, C4859 - Ready-made solution, lOOmg ml in DMSO, Sigma) was diluted in transparent growth medium to 10 μΜ. Normal transparent growth medium (2ml, without CHX) was replaced by the 2ml diluted CHX under the microscope after at least one round without the CHX. The dynamic profiles of the proteins were then measured for 20 hours (20 minutes resolution) using time-laps

SUBSTITUTE SHEET RULE 26 microscopy as described in previous sections. It was found that about 10 % of the cells underwent apoptosis after 20 hours as a result of CHX addition and most of the cells stopped dividing.

Measurements were done on 11 proteins (DDX5, CKS2, ENOl, JAGN1, DNCH1, RPS3, LMNA, MAP2K2, NCBP2, PSMB4 and RPL22) with at least two replicates. One usually assumes that the translation inhibitor CHX halts protein production and induces a time dependant exponential decay according to Eq. 11. However, the present inventors observed different behaviors: Most proteins decayed to a new steady state (instead of decaying to zero) and one protein, ENOl, showed an increase in response to CHX (Figure 2). About half of the tested protein tested showed dynamics that did not fit an exponential decay to zero (e.g. CKS2), suggesting that translation inhibition was incomplete. Increasing the dosage of CHX did not eliminate this effect (Figure 11).

Taken together, these measurements suggest that in the present human cells CHX does not halt translation but only reduces it (increasing CHX dosage by 10 fold gave similar protein profiles). Hence, retrieving degradation rates from the present measurements requires a model that allows for both production and degradation. The present inventors tested several models but found that the estimated degradation rate can change considerably (factor of 10) according to the choice of model. Overall, the present inventors were unable to obtain reliable degradation rates in the present human cells using a CHX based approach.

Bleach chase: theory and experimental protocol: This section describes how bleach-chase works by deriving Eq. 2 used for degradation rate measurement and the experimental protocol. Assume a fluorescently tagged protein whose degradation rate, , is constant in time (later we show how to retrieve degradation in case ^a changes in time). By exposing the cells to a brief pulse of light one can bleach a fraction of the proteins inside the cells effectively transforming them from fluorescently tagged into non-fluorescent. One can therefore think of the total protein, P , as the sum of two cohorts of proteins, one visible to fluorescent microscopy, Pv , and another which is invisible, P:

Eq. ilJ) Pit) = P„{t) + PitX The invisible protein P is produced only during the pulse, after which it starts to degrade according to the following equation:

Eq. (14) —— = -a _* Pit!

dt

Hence, measurement of how ^p changes in time would enable the retrieval of the degradation rate. But, since ^p is invisible it cannot be measured directly. The solution is to retrieve it indirectly by measuring ^p and ^pv and applying Eq. 13 and 14:

Eq. (15) ^-=-^ = -a * (F(¾ - ¾(t) ).

The solution to this equation means that the difference between the bleached and non-bleached experiment decays exponentially in time at a rate that depends solely on the degradation rate of the protein under study:

Eq. (P( - P_v{t)) = (P_Q - P_vBi e-^st.

and hence:

Eq.a.7 ln(p(£) - P_v(t)} = ln(P_Q(t) - P_vQ(t) aL

One can relax the assumption that the degradation rate is constant in time by applying instead of ^a in Eq. 17 and driving a new set of equations.

Protein fluorescence was measured with and without the drug in 12 well plates using multiple fields of view under time lapse microscopy as described in previous sections. However, in bleach-chase two concurrent measurements are done: without and with the bleaching. The latter was done using a pulse of light given one round prior to drug addition. In the rest of the time-lapse normal exposure times are used that are sufficient to detect the protein under study. We tested a wide range of pulse durations (1 - 8 minutes) using a mercury fluorescent lamp (120 W) resulting in different fractions of the protein fluorescence bleaching (10 % - 60 % drop). The measured degradation rates were roughly insensitive to the degree of bleaching (CV < 0.1).

A simple model: changes in growth-rate induce global changes in half-lives

In this section the present inventors derive Eq. 6, that captures the half-life after a change in growth rate, 2 , as a function of the half-life prior to the change, 2 , the cell cycle duration, ^cc , and the fold change in growth rate, & . By applying Eq. 3, it follows that:

ln(2)

Eq, {18)

2

ln(2)

2

ln(2)

Eq. (20) ¾ =

Applying Eq. 18 and 20 to Eq. 19 gives:

T *

«d_eg+«dii+(k - 1)^■ <¾n ^fa (²) ₊ (jfc - i) ¹¹¹ (²>

¹1/2 ⁱ C and hence we get Eq. 6:

EXAMPLE 1

Parallel measurement of Protein and mRNA dynamics in human cancer cells Results

The dynamics of 540 proteins and their mRNAs were followed for 48 hours as cancer cells responded to the drug CPT (experiment summarized in Figure 3). These genes represent a wide range of biological functions and processes (Table 4).

Table 4

Entrez

Gene Description

23 ATP-binding cassette, sub-family F (GCN20), member 1

11332 acyl-CoA thioesterase 7

71 actin, gamma 1

81 actinin, alpha 4

10121 ARP1 actin-related protein 1 homolog A, centractin alpha (yeast)

cacum c anne, votage- epen ent, ap a eta su unt

cleavage stimulation factor, 3 pre-RNA, subunit 3, 77kDa cancer/testis antigen IB

cullin 3

death-associated protein

DAZ associated protein 2

discoidin, CUB and LCCL domain containing 2

development and differentiation enhancing factor 2

DEAD (Asp-Glu-Ala-Asp) box polypeptide 18

DEAD (Asp-Glu-Ala-Asp) box polypeptide 21

DEAD (Asp-Glu-Ala-Asp) box polypeptide 46

DEAD (Asp-Glu-Ala-Asp) box polypeptide 5

DEK oncogene (DNA binding)

DEAH (Asp-Glu-Ala-His) box polypeptide 15

dickkopf-like 1 (soggy)

discs, large (Drosophila) homolog-associated protein 1

cyclin D binding myb-like transcription factor 1

DnaJ (Hsp40) homolog, subfamily A, member 1

DnaJ (Hsp40) homolog, subfamily C, member 9

DNA (cyiosine-5-)-mefhyltransferase 1

dihydropyrimidinase-like 3

Down syndrome critical region gene 2

destrin (actin depolymerizing factor)

dual specificity phosphatase 18

dynein, cytoplasmic 1, heavy chain 1

dynein, light chain, roadblock-type 1

eukaryotic translation elongation factor 1 epsilon 1

eukaryotic translation elongation factor 1 gamma

eukaryotic translation elongation factor 2

eukaryotic translation initiation factor 1A, X-linked

eukaryotic translation initiation factor 2, subunit 3 gamma, 52kDa eukaryotic translation initiation factor 3, subunit A

eukaryotic translation initiation factor 4A, isoform 1

eukaryotic translation initiation factor 4E

eukaryotic translation initiation factor 4E family member 2

eukaryotic translation initiation factor 4E binding protein 1

eukaryotic translation initiation factor 4 gamma, 3

eukaryotic translation initiation factor 4H

eukaryotic translation initiation factor 5B

epithelial membrane protein 3

enolase 1, (alpha)

endosulfine alpha

enhancer of rudimentary homolog (Drosophila)

electron-iransfer-flavoprotein, beta polypeptide

family with sequence similarity 128, member A

family with sequence similarity 44, member A

3151 high-mobility group nucleosomal binding domain 2

51335 neugrin, neurite outgrowth associated (NGRN), transcript variant 1,

4 paraoxonase

54922 Ras interacting protein 1 1

3

1

6636 small nuclear ribonucleoprotein polypeptide F

6637 small nuclear ribonucleoprotein polypeptide G

10073 snurportin 1

8724 sorting nexin 3

58533 sorting nexin 6

6647 superoxide dismutase 1, soluble (amyotrophic lateral sclerosis 1

(adult))

65244 spermatogenesis associated, serine-rich 2

28972 signal peptidase complex subunit 1 homolog (S. cerevisiae)

23111 spastic paraplegia 20 (Troyer syndrome)

6711 spectrin, beta, non-erythrocytic 1

144108 SPT2, Suppressor of Ty, domain containing 1 (S. cerevisiae)

23350 U2-associated SR140 protein

10250 serine/arginine repetitive matrix 1

23524 serine/arginine repetitive matrix 2

10274 stromal antigen 1

10963 stress-induced-phosphoprotein 1 (Hsp70/Hsp90-organizing protein)

6789 serine/threonine kinase 4

3925 stathmin 1/oncoprotein 18

7341 SMT3 suppressor of mif two 3 homolog 1 (S. cerevisiae)

6613 SMT3 suppressor of mif two 3 homolog 2 (S. cerevisiae)

8887 Taxi (human T-cell leukemia virus type I) binding protein 1

64786 TBC1 domain family, member 15

1155 tubulin folding cofactor B

22980 transcription factor 25 (basic helix-loop-helix)

6950 t-complex 1

6996 thymine-DNA glycosylase

26136 testis derived transcript (3 LIM domains)

7019 transcription factor A, mitochondrial

29844 TCF3 (E2A) fusion partner (in childhood Leukemia)

80764 THAP domain containing 7

10189 THO complex 4

7077 TIMP metallopeptidase inhibitor 2

7082 tight junction protein 1 (zona occludens 1)

116238 TLC domain containing 1

55002 transmembrane and coiled-coil domains 3

114908 transmembrane protein 123

121256 transmembrane protein 132D (TMEM132D), mRNA

81671 transmembrane protein 49

7112 thymopoietin

7125 troponin C type 2 (fast)

54543 translocase of outer mitochondrial membrane 7 homolog (yeast)

9868 translocase of outer mitochondrial membrane 70 homolog A (S.

cerevisiae)

7150

7168

7170

7171

1200

29896

79090

7706

203062

79989

10376

84790

81027

10383

7295

9352

7296

10587

7307

6675

7322

7332

7334

92912

84993

7357

7360

26019

7381

7386

9100

219333

9097

51118

7402

9218

7408

7411

7414

7430

7431

9559

9525

7453 tryptophanyl-tRNA synthetase

55759 WD repeat domain 12

144406 WD repeat domain 66

339005 WAS protein homology region 2 domain containing 1-like 1

9503 X antigen family, member ID

7520 [1x125 char]

2547 [1x94 char]

10138 YY1 associated factor 2

10413 Yes-associated protein 1, 65kDa

4904 Y box binding protein 1

91746 YTH domain containing 1

7531 [1x91 char]

7534 [1x88 char]

7528 YY1 transcription factor

9839 zinc finger E-box binding homeobox 2

51663 zinc finger RNA binding protein

23613 zinc finger, MYND-type containing 8

9726 zinc finger protein 646

7784 zona pellucida glycoprotein 3 (sperm receptor)

9183 ZW10, kinetochore associated, homolog (Drosophila)

The proteome response was measured using a dynamic proteomics approach [14, 16-18]. A library of human lung cancer clones (H1299 cell line) was used, where each clone has a different endogenous protein fused to a yellow fluorescent protein (YFP) that is used as a reporter. The YFP was inserted as an artificial exon at the natural chromosomal loci of the gene, resulting in a full length protein that is under its native regulation. Comparisons to immunoblots indicated that 80 % of the tagged proteins are accurate markers for the endogenous protein dynamics [18]. All the clones in the library contain an additional red fluorescent marker that enables accurate automated image analysis and tracking of individual cells. Analysis was performed automatically using custom software (Methods and Figure 4)

Protein dynamics were measured at high temporal resolution (every 20 minutes, 144 time points) using time lapse microscopy under incubated conditions in multi-well plates as described [18]. Experiments were done in two repetitions on average with four fields of views taken from each well. For each protein, the dynamics were obtained by automatically tracking 300 - 600 individual cells and averaging their profiles. After 24 h of growth, the topoisomerase-l poison CPT [21] was added at 10 μΜ. All cells stopped dividing 5-7 hours after drug addition and about 25 % underwent apoptosis after 48 hours. Assay reproducibility in day to day repeats was better than 20 % (std/mean). Thus changes larger than 20 % in a tagged protein intensity can be reliably detected using the present assay.

Dynamic proteomics has previously been used to measure the temporal changes of proteins compared to a reference time-point. Here dynamic proteomics was extended to enable comparisons across different proteins. For this purpose a calibration was developed that converts protein fluorescent units into relative protein abundance. The calibration employs the red fluorescence marker common to all the clones as a reference for normalizing the different proteins (details in Methods). This calibration yielded excellent agreement with quantitative immunoblots on selected proteins.

mRNA dynamics was measured using Affymetrix GeneChip Human Gene 1.0 ST arrays that comprehensively cover known human transcripts. Cells were grown in 10 cm plates in the same conditions as the protein experiments and harvested at nine different time points before and after drug addition (0, 1.5, 3, 6, 9, 12, 23, 33 and 48 hours) with at least three replicates per time point. Data was normalized using standard RMA background correction and GC content normalization. Reproducibility was better than 15 % (std/mean) (Figures 5A-B). To further test the microarray measurements, qPCR was performed on 10 genes that showed diverse dynamics. The measurements obtained by the two assays yielded very similar profiles with a median correlation of 0.83 (Figure 6). Time-dependant gene ontology (GO) enrichment analysis revealed temporal waves of functionally related genes that responded in concert to the drug: some falling together (e.g. histone and nucleosome assembly constituents P < 10^"60, cell division related genes P < 10^"13 and snoRNAs P < 10^"15) and other rising (e.g. MHC-I genes P < 10^"8 and lysosomal genes P < 10^"8) (for details see Figures 1A-K).

Finally, the protein and mRNA sets were intersected to create a dataset of 540 protein and mRNA pairs of co-dynamic responses (Table 4, Figure 7). EXAMPLE 2

Dis-correlation between dynamics ofmRNAs and their corresponding proteins in response to a drug

The present inventors asked whether changes over time in a given protein are related to the changes in that protein's mRNA levels. To examine this the temporal dynamics of 540 proteins were compared to their corresponding mRNAs. This revealed a wide range of co-dynamic behaviors. About 30% of the mRNA-protein pairs exhibited strong positive correlations in their temporal profiles (p > 0.6). Strikingly, strong anti- correlations (p < -0.6) were found in 15 % of the genes in response to the drug. The rest of the genes (about 55 %) showed weak or no correlation to their mRNA (-0.6 < p < 0.6) (see Figures 8A-B).

The significance of the observed correlations was tested _<by a data shuffling test, in which the identities of mRNAs and proteins were randomly shuffled. It was found that the dynamics of only about 15 % of the proteins exhibited a positive correlation that is more than expected in shuffled data (P < 10^"6, Figure 2). This suggests that under the present conditions only a small fraction of the mRNA variations over time determine the variations in protein levels.

Time-shift correlation analysis was also performed, in which the temporal protein profiles were shifted with respect to the mRNA profiles, to test for effects of delay between mRNA and protein. It was found that time shift did not yield more positive correlations than observed in shuffled data (P ~ 0.5). This suggests that time delay in protein production is not a major factor causing the observed dis-correlation.

It was found that sets of genes sharing similar functions tended to share similar correlations or lack of correlations between their mRNA and protein dynamics. For example, ribosomal proteins were dis-correlated to their mRNAs in a systematic manner: whereas most ribosomal proteins decay in response to the drug (GO enrichment P < 10^"4) their corresponding mRNAs all showed a synchronized increase in the first 8 hours and then a slow decrease (Figures 8C-E). In contrast, most cytoskeleton related proteins showed a correlated decay in both protein and mRNA that was synchronized with loss of cell motility (Figures 8F-H).

In addition to following the temporal correlation between pairs of mRNA and protein over time, the correlation across different genes using snapshots of the transcriptome and proteome at single time points was also tested. The present data has a dynamic range of about 500 fold between the least and most abundant protein. The present inventors asked whether these differences in protein levels are compatible with variations in their mRNAs levels. It was found that relative mRNA levels do not correlate well with relative protein levels across different genes (r = 0.13, P < 10^"9), Figure 21. Correlation remained roughly constant in snapshots taken at different times after drug addition (r = 0.1 - 0.15). This result is in agreement with previous studies that compared mRNA and protein snapshots in human cells [8-13].

Overall, it was found that snapshots as well as time dependant changes in pairs of proteins and mRNAs are uncorrelated and often anti-correlated in human cancer cells under the present condition. This suggests that post-transcriptional control is a key factor in determining protein dynamics in response to the drug.

The present inventors next asked how protein translation and degradation rates combine to produce the observed dynamics. It was found that kinetic models with constant translation and degradation rates poorly fitted the data, whereas models with time varying translation and degradation were not constrained enough to explain the data: changes in either production or degradation could result in the same dynamic profile of protein levels. Hence, to proceed, direct measurements of protein dynamic degradation and translation rates were measured.

EXAMPLE 3

Bleach-chase: a novel method for accurate measurements of

protein degradation rate dynamics

Protein dynamics is dictated by production of new proteins and their removal. The latter is the sum of two underlying processes: intra cellular degradation (e.g. due to proteosome activity) and dilution due to cell growth that effectively reduces the protein amount by 50% every one cell generation [22]. For clarity the term degradation rate, a, will refer to the sum effect of both intra cellular degradation, ^_eg, and the dilution rate, ctdii :

Eq. (1) a = a_{deg +} a_dii Then protein levels, then follows the simple equation [22]:

, dp

Eq. ^■= βίύ - ait)^■ Pit) .

where jS(t} is the production rate and «x(t) the degradation rate (note that production and degradation may also implicitly depend on mRNA and protein levels). The present goal in this section is to measure the production and degradation rates fi(t) and (t) as a function of time.

The present inventors first attempted to measure protein degradation rates using a standard translation inhibition protocol [23]. In this approach protein levels are measured at different times after the addition of the translation inhibitor Cyclohexamide (CHX). It is expected that protein production will halt and that the protein dynamics will gradually decay to zero in a rate that is exponential in the degradation rate. The present inventors followed this protocol and added the inhibitor CHX, with and without the drug CPT, at different concentrations and measured the levels of 11 proteins with our dynamic proteomics assay (see Methods). This method did not yield satisfactory results. The present inventors observed one protein that increased its level rather than decreased. Furthermore, about half of the proteins tested showed dynamics that did not fit an exponential decay to zero levels, suggesting that translation inhibition was incomplete (increasing the dosage of CHX did not eliminate this effect) (Figure 2). Additionally, about 10 % of the cells underwent cell-death and most cell stopped dividing due to the CHX addition. Therefore, extracting meaningful degradation rates for many of the tested proteins was not possible using this approach.

An alternative method for measuring protein degradation rates is 'pulse-chase' in which proteins are radioactively labeled over a brief period (the pulse) and then the decay in radioactivity is measured in time (the chase) and used to infer protein half life. This assay requires a different antibody for each protein of interest in order to pull it down. Despite its accuracy, it is difficult to scale this method up to more than a few proteins and time points.

In order to address these limitations and measure protein degradation and production rates of multiple proteins at high temporal resolution without the need for translation inhibitors or radio-active labeling, the present inventors developed an assay called 'bleach-chase'. In bleach-chase, one bleaches the fluorophore of the tagged protein using a brief pulse of light thus irreversibly turning it non-fluorescent (it is sufficient to bleach only a small fraction of the tagged proteins). Following the cells over time shows that protein fluorescence recovers over hours to days, due to production of new tagged proteins. The bleaching did not alter cell-viability, motility, mitosis rate or morphology. This is in agreement with the general observation that GFP- fused proteins are relatively non-phototoxic under fluorescence microscopy and retain protein activity under mild bleaching [24-25]. This method has been termed 'bleach- chase', to highlight the analogy with the classical 'pulse-chase' method. In bleach-chase cells are subjected to a mild pulse of light, which transforms a fraction of their fluorescent proteins into non-fluorescent. The decay of the non-fluorescent protein fraction (the difference between the fluorescence of the bleached and unbleached cells) is then chased in time to reveal the protein half-life: this difference decays in time according to the degradation rate, as described by the equation (obtained by subtracting and solving Eq. 1 for the bleached and unbleached cells, see Methods):

Eq (3) /n(P(t)-P_v(t)) = /n(P₀-Pov) - . t,

where P(t) and P_v(t) are the fluorescence levels of the unbleached and bleached cells, respectively; P₀ and P₀v are the fluorescence levels of the unbleached and bleached at an initial time point, respectively. If the degradation rate, a, varies with time t

the same equation applies, but a is replaced with ₀i ct(t') dt'. The formula is derived in the Methods.

To understand this method, note that the profiles of the bleached and unbleached cells begin at different initial levels and eventually converge to the same dynamics (Figures 9A-C). The rate at which they converge reveals the degradation rate: rapid convergence means rapid degradation and slow convergence means slow degradation. It is important to note that bleach-chase applies regardless of how protein production rate changes in time, nor does it require halting production. The intuition behind this result is that the production of the bleached and non-bleached cells is the same (bleaching does not eliminate the protein, only inactivates its tag making it invisible to fluorescent microscopy) and thus one can cancel-out the effects of production by subtracting the two dynamics. The day to day error of bleach-chase is about 0.4 (std/mean). Hence, the present inventors averaged of over 3 - 4 replicates to obtain accurate measurements. The present inventors also performed radioactive pulse-chase experiments and compared the measurements to those obtained using bleach-chase. The two methods yielded highly similar measurements with average deviation less than 15% (see Table 5, herein below).

Table 5

The present inventors also compared the degradation rate of the YFP-tagged protein to the non-tagged endogenous protein expressed form the untagged allele in the same cells and found them to be very similar (average deviation less than 10 %, Table 6)· Table 6

This result is in line with previous studies that found good agreement between half-lives and dynamics of fluorescently tagged endogenous proteins and untagged ones. As a final test of the bleach-chase method, bleaching at different intensities was applied. The measured degradation rates remained similar (CV < 0.2), regardless of bleaching intensity (Figure 9C). These tests suggest that bleach-chase is an accurate method for estimating protein degradation rates.

EXAMPLE 4

Protein half-lives increase in response to the drug

In this example, bleach chase is used to investigate the reasons for the dis- correlation between transcript and protein dynamics. To address this, 20 proteins were selected spanning different cellular localizations and functions bleach-chase was used to measure their degradation rates 0 - 48 hours after drug addition at high temporal resolution (every 10 minutes) with 4 - 6 repeats. The degradation rates in cells were cultured without the drug (4 - 6 repeats) was measured and the results compared. Strikingly, a global shift in protein degradation rates was observed due to the drug: most proteins either decreased their degradation (14 out of 20) or retained the same level (5 out of 20) as depicted in Figure 10A.

Overall, it was found that upon drug addition, degradation rates drop and then retain roughly constant levels (Figure 10B).

EXAMPLE 5

The longer the protein half-life, the larger the increase due to growth arrest

This example investigates possible mechanisms that may explain the global effect described in Example 4.

A protein's half-life, ^, given a constant degradation rate, a, is given by:

In (2)

Therefore, the observed global decrease in degradation rates is tantamount to an increase in protein half-lives. The present inventors found that the increase showed the following trend: the larger the protein half-life, the larger its increase after drug addition. In other words, stable proteins become increasingly more stable whereas unstable proteins remain largely unaffected by the drug. This phenomenon is shown in Figure IOC.

One plausible explanation for the observed systematic increase in protein half- lives is a global down-regulation of the protein degradation machinery (e.g. inhibition of ubiquitin-proteosome mediated proteolysis). However, this explanation does not account for the observation that stable proteins become increasingly more stable. On the contrary, stable proteins are expected to be the least affected, since their half-lives are largely determined by dilution due to cell-division rather than intra-cellular degradation.

Next, the present inventors turned to focus on an alternative simple model that accounts for the observed global half-life behavior. It assumes that the drug induces cellular growth arrest. This assumption was tested by obtaining quantitative measurements of the mitosis rates at different time points before and after drug addition using the image analysis software. It was found that mitosis rate drops upon drug addition and reaches a complete stop after 5 - 7 hours.

To see why growth arrest is sufficient to produce the observed half-live behavior, one must return to Eq. 1 that describes protein degradation as the sum of intracellular degradation and dilution due to cell growth. A decrease in growth rate thus affects the degradation rates of all proteins as follows:

Eq. (5) a = a_{deg +} k. a_dii

where 0 < k < 1 defines the degree of growth arrest. Note, that although the decrease in dilution rate is the same across all proteins, the relative change in a_tot may vary considerably. Unstable proteins have rapid intra-cellular degradation (a^g » α^ι) and thus their a_tot remains unaffected even when dilution rate drops to zero. This is in agreement with our observation that proteins with short half-lives are less affected by growth arrest. Stable proteins, on the other hand, have slow intra-cellular degradation (ct_deg « their half life is mainly determined by the rate of cell division, and hence its arrest is expected to significantly increase protein half-life. This is also in agreement with our measurements.

Next, the present inventors tested whether the growth arrest model can quantitatively explain the measured changes in half life. One can predict a protein's half-life after growth arrest, T^₂, based on its half-life prior to the arrest, T_1/2, and the average cell cycle time, T_m using the following equation (based on Eq. 3 and 4, see Methods): *fl. (6) ¾ = ^T T , >

cc

The human lung cancer cells studied here have an average cell cycle time of 20 hours (Tec ⁼ 20 + 3 hours). Applying this information with the observation that cell cycle stops after the drug addition (k = 0) to Eq. 6 yields the predicted half-life behavior depicted by the blue line in Figure IOC (R² = 0.4). Deviations from the line are due to protein specific regulation.

One interesting property that emerges from the model is that the increase in protein stability due to growth arrest is inversely proportional to the cell cycle duration, which means that the faster cells divide prior to the drug addition, the stronger the affect as illustrated in Figure 10D.

Overall, it was found that protein lifetime increased after drug addition, an effect that is stronger the longer the pre-drug lifetime of the protein is. The present analysis suggests that this effect is largely due to cell-growth arrest. EXAMPLE 6

Protein translation gradually decreases in response to the drug Next, the dynamics of the translation rates in response to the drug were tested.

To this end the production rate fi^ in Eq. 1 was estimated, which accounts for both transcriptional and post transcriptional effects influencing protein production. The production rate for these proteins can be determined using the measured degradation dynamics (i.e. applying Eq. 1). The translation rate dynamics, Y^, is then determined using the production rate and the measured mRNA dynamics, according to the following equation:

£¾. (7) β® = γΙ · Μίύ.

Note, that F© accounts for all post transcriptional effects influencing protein production.

It was found that for most of the proteins, translation decreases gradually to less than half of the initial level (Figure 11 A, B). The decay in translation is similar in timing for most proteins in this study, suggesting a common cause. This is in line with the observation that CPT causes a decrease in ribosomal proteins on a similar timescale (Figure 11C, p = 0.8). These results taken together with the half-life measurements indicate that the present drug has two opposing effects: a global increase in protein half- lives and simultaneous decrease in translation rates. The balance between these counter effects significantly changes in response to the drug, a change which varies considerably across different proteins, as depicted in Figure 11D. The change in balance between half-lives and translation rates, rather than changes in mRNA dynamics, is in many instances the dominant factor controlling the protein response (Figure 12). Figure HE, shows examples of proteins whose temporal response is dis-correlated and even anti-correlated to their mRNAs dynamics but in accordance with their relative increase in half-lives.

EXAMPLE 7

Additional information

1. A kinetic model with constant translation and degradation rates does not explain the data

It was asked whether a simple kinetic model that assumes constant translation and degradation rates may explain the observed dis-correlation between mRNA and protein. To address this, the present inventors used the following model:

where Y is the translation rate (note that Y accounts for all post-transcriptional production), « is the degradation rate, P(t) is protein levels and M(t) mRNA levels. Both M(t) and P(t) were normalized to their levels at t = 0 prior to drug addition. M(t) was then interpolated to match the sampling frequency of P(t) yielding two vectors with 144 time points. Hence, for each pair of mRNA and protein profiles an over-determined linear set of equations (143 equations and two parameters) is obtained. Translation and degradation rates were then estimated by identifying the two values that minimize the sum of square errors over all equations (other estimation procedures such as minimization on the integral solution yielded similar results). One can then use the estimations of translation and degradation rates together with mRNA measurements to predict protein levels using the integral solution of Eq. SI:

Eq. (52) Pit) = P₀e-** + J γ · M{k)e^^~kydk

o

Next, the predicted protein profiles were compared to the empirical protein measurements and found that in about half of the cases the difference between the predicted and empirical protein levels was greater than the present experimental measurement error (CV < 20 %) indicating that the above model is inadequate for capturing the dynamics of many of the proteins. This suggested that the model assumption, that translation and degradation are roughly constant in response to the drug, is incorrect and that at least one of these components is significantly changing in time.

2. Kinetic models with time varying translation and degradation rates yield non- unique predictions

The present inventors therefore, turned to investigate models that allow for temporal changes in the behavior of the translation and degradation. To this end the following more general model was used: dPtO

Eq. (S3) —— = yit ^■ Mit) - am^■ P(t

at

Note that since this model allows both translation and production to change in time, there are more parameters than equations. To circumvent this problem of over- fitting two approaches were tested: The first was to penalize for non-smooth profiles of y(t) and «(t) thus restricting the rate of temporal change. The second was to use a piecewise linear approach where both translation and degradation profiles are modeled using 2 - 5 connected lines. While both methods were able to accurately capture most protein behaviors (i.e. error of fit was better than measurement error) the parameter estimations were non-unique: different profiles of production or degradation could result in the same dynamic profile of protein levels. Overall, the present inventors were unable to indirectly retrieve reliable translation and degradation rates from mRNA and protein dynamics alone. We therefore turned to direct measurements of protein degradation and translation rate dynamics.

3. A decrease in intra-cellular degradation is expected to induce a global change in half-lives

This section discusses a simple model for predicting the shift in half-lives due to a global down-regulation of intra-cellular degradation (e.g. inhibition of ubiquitin- proteosome mediated proteolysis). This model suggests that the half-lives of all proteins is expected to increase but that stable proteins will be the least affected. The rational is that their half-lives are largely determined by cell-division rather than intra-cellular degradation.

First, the model assumes that the process of protein degradation is a sum of two underlying processes: (i) intra-cellular active degradation, ot^, (e.g. due to proteosome activity) and (ii) protein dilution rate due to cell growth, ^adu, which effectively reduces the protein amount by 50 % after one cell generation time. A protein's effective degradation rate, &tot , can therefore be captured by the following equation: Eq. (S4) a = a_d¾g + a_mi

A global inactivation of intra-cellular degradation can be assumed (due to down regulation of one of the essential components) that can be captured according to the following equation:

Eq. (55) = k - _deg + a_aa where 0— ≤ ¹ defines the degree of reduction in intra-cellular degradation. The goal now is to derive the new protein half-life after, 2 , based on its half- rl

life prior to the inactivation, 2 , and the average cell cycle time, * cc . Using Eq. 3, S4 and S5 one can derive the following relations:

£¾, (56) 7i =

Ε (57) Γί = M2)

Eq. (S8) T_cc =

Applying Eq. S6 and S8 to S7 gives: ln(2) ln(2)

H^adeg + «dil) + C¹ - K)^aM

¹cc resulting in:

This model suggests that when intra-cellular degradation decreases the result is a global increase in half lives but stable proteins are expected to be the least affected. This point is illustrated in Figure 14, where the predicted increase for different k's are shown given a T∞= 20 hours (as measured in the cells used in this study).

Discussion

This study provides an integrated view of the dynamics of the transcriptome, proteome, protein degradation and translation rates as human cancer cells respond in time to a drug. The present inventors asked how well are temporal changes in the proteome correlated with changes in the transcriptome. Measurements of 540 protein dynamics and their respective mRNAs in response to the drug revealed poor correlations and even anti-correlations in 70 % of the pairs. Time shift correlations for capturing potential delays between mRNA and protein as well as kinetic models did not resolve the observed discrepancy, suggesting that post-transcriptional control plays an important role in fine tuning the response to the drug CPT. Whether the temporal dis- correlation between mRNA and protein observed in the present condition applies to other non-drug like stimuli in human cells is still an open question.

Next, the present inventors asked how temporal changes in post transcriptional control, namely protein translation and degradation, combine to produce the observed protein dynamics. Current methods for measuring protein degradation rates have several limitations: pulse-chase is accurate but requires radioactive labeling, protein specific antibodies and is laborious, limiting its usage to the measurements of few proteins at a time; the translation inhibitor CHX is non-radioactive and can be used to measure multiple proteins but is highly pertubative to the cells and may alter the native degradation rates due to translation inhibition of ubiquitine related proteins. To overcome these limitations the present inventors have devised a novel method for accurate measurements of protein half-lives in living cells - bleach-chase. It is non- radioactive, simple to apply, and can be readily scaled to measure multiple proteins at high temporal resolution. The mild pulse of bleaching is also significantly less pertubative than translation inhibition. The main limitation of bleach-chase is the requirement that the protein under study be fluorescently tagged. While, tagged proteins tend to preserve the same dynamics and half-lives as untagged ones (Table 6) their availability is still limited to a few organisms (e.g. the yeast GFP tagged protein library [29] and the human LARC library [18]). Broad usage of bleach-chase will become more feasible with the expected generation of additional tagged protein libraries in other organisms and cell-lines. One future application of bleach-chase is to study how localization changes affect protein half-life dynamics in living cells.

The present inventors used bleach-chase to measure protein degradation dynamics in response to the drug. A global increase in the half-lives of most proteins was observed. Notably, the larger the protein half-life prior to drug addition, the larger the increase in half-life after addition (Figure IOC). To explain the global increase in life-times, a simple quantitative model may be proposed, which assumes that the drug induces cellular growth arrest. It was found that cells stop dividing after drug addition and that the model accurately reproduces the observed trend. Several properties emerge from this model. First, it suggests that the increase in life time may be a general property of cells that slow down their growth. Second, it predicts that the faster the growth prior to growth arrest, the steeper the increase in half -lives. This implies that rapidly dividing cells (e.g. cancer cells or hair follicles) will experience a larger disequilibrium in their half-lives compared to slowly dividing cells (e.g. neurons) as a result growth arrest (Figure 10D) thus creating an imbalance in protein levels in the cell, with stable proteins rising in levels more than unstable ones. This might be related to the ability of growth arrest drugs to differentially harm fast dividing cells such as tumor cells. Third, an increase in cell-growth is expected to induce the opposite effect: a global decrease in half-lives in which stable proteins undergo a large decrease and unstable proteins remain largely unaffected. These predictions are now readably testable in multiple conditions and proteins using bleach-chase.

Finally, the temporal changes in translation rates were determined in response to the drug. A gradual decrease in translation rates was observed, whose timing agrees with the decrease of ribosomal proteins. Taken together with the measurements of half- lives, it was found that the drug induces two global, opposing effects: a decrease in translation and an increase in half-lives. The balance of these effects varies widely across different proteins. This balance was found to be a dominant factor in explaining protein dynamics in response to the drug resolving the observed dis-correlation between the proteome and transcriptome. The present approach opens the way for a global study of the time-dependant interplay between the transcriptome, translation and protein degradation rates and how they shape the dynamic response of the proteome to different stimuli.

EXAMPLE 8

Further studies using Bleach chase .

The present inventors compared the bleach-chase method for analyzing half- lives as described herein with radioactive pulse-chase experiments, the gold standard, on additional proteins (above those analyzed in Example 3) and showed similar half- lives for tagged proteins (11% median difference, seven proteins) as illustrated in Table 7 and Figures 15A-B.

The difference between the two methods was assessed using the following formula. Average difference was 10 %. ipblemch— chase J"'^m' Ise— cha e ipbleaeh—chase y>P w¹ ise— chase

Table 7

The present inventors compared additional tagged proteins (above those analyzed in Example 3) to their untagged counterparts (16% median difference, six proteins), as illustrated in Table 8 and 015A-B.

In 5 out of 6 cases it was found that the YFP-tagged and its corresponding native protein showed similar half-lives (average difference of 14%). ^ , _{; ||}_ _2011/ .₀6-2011

68

Table 8

* imtnunoblots showed a missing 17 KD in the YFP tagged protein possibly due to a

truncation caused by the tagging. This may account for the observed discrepancy

between the half-life of tagged and untagged STMN1.

As a further test of the method, bleaching was applied at different intensities.

The measured removal rates remained similar (CV<0.2), regardless of bleaching

intensity (Figures 16A-B). Thus, bleach-chase seems to accurately measure protein

removal rates in living cells.

The present inventors next asked which process, degradation or dilution,

dominates protein removal. Initial experiments were conducted using growth

conditions in which the cells vigorously divided, and bleach-chase was used to assay

100 proteins spanning different cellular localizations and functions for 24 hours

(every 20 minutes) with 3-4 day-to-day repeats. A broad protein half-life distribution

was observed ranging between 45 minutes to 22.5 hours, with mean 9.0±4.6h (Figure

17B and Table 9, herein below).

Table 9

CALM2 calmodulin 2 6.7 + 1.1

CD44 CD44 antigen isoform 1 precursor 20.8 + 1.9

CIRBP cold inducible RNA binding protein 6.9 + 2.1

CKS2 CDC28 protein kinase 2 2.6 + 1.6

COPS6 COP9 signalosome subunit 6 5.4 + 1.0

COTL1 coactcsin-like 1 8.4 + 1.4

COX7C cytochrome c oxidase subunit VIIc precursor 22.5 + 5.1

DDX18 DEAD (Asp-Glu-Ala-Asp) box polypeptide 18 20.3 + 1.4

DDX46 DEAD (Asp-Glu-Ala-Asp) box polypeptide 46 7.9 + 2.2

DDX5 DEAD (Asp-Glu-Ala-Asp) box polypeptide 5 8.7 + 1.0

DNMT1 DNA (cytosine-5-)-methyltransferase 1 6.9 + 2.0

DYNC1H1 dynein, cytoplasmic, heavy polypeptide 1 13.8 + 4.3

EEF1A1 eukaryotic translation elongation factor 1 alpha 9.9 + 0.6

EEF1E1 eukaryotic translation elongation factor 1 2.7 + 0.7

EEF2 eukaryotic translation elongation factor 2 8.8 + 0.7

EIF2S2 eukaryotic translation initiation factor 2 beta 8.3 + 1.5

EIF2S3 eukaryotic translation initiation factor 2, 10.5 + 2.5

EIF4A1 eukaryotic translation initiation factor 4 A 7.3 + 1.2

EIF4E eukaryotic translation initiation factor 4E 11.7 + 0.9

EIF4EBP1 eukaryotic translation initiation factor 4E 8.5 + 2.3

EIF4G3 eukaryotic translation initiation factor 4 8.8 + 1.1

EIF4H eukaryotic translation initiation factor 4H 16.3 + 6.2

EIF5B eukaryotic translation initiation factor 5B 8.9 + 2.8

ENOl enolase 1 8.0 + 3.5

FAU ubiquitin-like protein fubi and ribosomal 12.2 + 1.2

FSCN1 fascin 1 5.8 + 0.2

GAPDH glyceraldehyde-3-phosphate dehydrogenase 0.8 + 0.8

H2AFV H2A histone family, member V isoform 2 13.9 + 1.8

HAT1 histone acetyltransferase 1 isoform a 8.5 + 1.2

HDAC2 histone deacetylase 2 4.5 + 2.4

HMGA1 high mobility group AT-hook 1 isoform a 11.1 + 1.5

HSP90AA1 heat shock protein 90kDa alpha (cytosolic), 4.2 + 1.8

HSP90AB1 heat shock 90kDa protein 1, beta 8.8 + 2.4

HSPA4L heat shock 70kDa protein 4-like 10.6 + 2.7

HSPH1 heat shock 105kD 16.5 + 4.2

ILF2 interleukin enhancer binding factor 2 16.5 + 3.5

K-a-1 tubulin, alpha lb 15.3 + 2.4

LMNA lamin A/C isoform 2 12.8 + 2.2

MYH9 myosin, heavy polypeptide 9, non-muscle 6.4 + 0.6

NASP nuclear autoantigenic sperm protein isoform 1 16.3 + 2.3

NCL nucleolin 13.9 + 2.8

NDUFAF2 NADH dehydrogenase (ubiquinone) 1 alpha 6.6 + 1.4 NDUFB11 NADH dehydrogenase (ubiquinone) 1 beta 10.8 + 1.2

NPM1 nucleophosmin 1 isoform 1 5.2 + 0.1

PLEC1 plectin 1 isoform 6 11.7 + 4.6

POLR2F DNA directed RNA polymerase II polypeptide F 5.4 + 1.8

POLR2L DNA directed RNA polymerase II polypeptide L 10.5 + 2.6 polymerase (RNA) III (DNA directed)

POLR3GL polypeptide 3.5 + 0.0

POMP proteasome maturation protein 7.0 + 2.9

PRDX5 peroxiredoxin 5 precursor, isoform a 9.4 + 2.6

PSMA1 proteasome alpha 1 subunit isoform 2 8.8 + 3.1

PSMA3 proteasome alpha 3 subunit isoform 1 4.0 + 4.4

PSMA7 proteasome alpha 7 subunit 9.0 + 2.6

PSMB1 proteasome beta 1 subunit 6.7 + 0.1

PSMB4 proteasome beta 4 subunit 7.3 + 0.8

PSMB6 proteasome beta 6 subunit 8.5 + 4.7

PSMB7 proteasome beta 7 subunit proprotein 5.1 + 0.7

PSMC1 proteasome 26S ATPase subunit 1 5.2 + 1.5

PSMC4 proteasome 26S ATPase subunit 4 isoform 1 4.4 + 1.2

PSMD12 proteasome 26S non-ATPase subunit 12 isoform 1 4.0 + 2.0

RFC1 replication factor C large subunit 9.3 + 1.5

RPA2 replication protein A2, 32kDa 11.2 + 2.0

RPL11 ribosomal protein Lll 14.9 + 2.3

RPL18 ribosomal protein LI 8 9.2 + 2.0

RPL22 ribosomal protein L22 proprotein 11.1 + 2.8

RPL27 ribosomal protein L27 15.4 + 3.1

RPL27A ribosomal protein L27a 4.2 + 1.9

RPL29 ribosomal protein L29 4.8 + 1.8

RPL30 ribosomal protein L30 6.1 + 0.6

RPL39 ribosomal protein L39 3.4 + 0.5

RPL6 ribosomal protein L6 9.3 + 1.6

RPS19BP1 S19 binding protein 21.8 + 0.9

RPS2 ribosomal protein S2 19.7 + 3.5

RPS3 ribosomal protein S3 5.5 + 2.4

RPS3A ribosomal protein S3a 6.7 + 1.0

RPS6 ribosomal protein S6 14.4 + 3.5

RPS7 ribosomal protein S7 4.3 + 1.3

RPS8 ribosomal protein S8 3.0 + 0.2

SLBP histone stem-loop binding protein 8.2 + 1.1

SSBP1 single-stranded DNA binding protein 1 6.9 + 0.3

STMN1 stathmin 1 3.9 + 0.6

TARS threonyl-tRNA synthetase 6.6 + 1.1

TBCA tubulin-specific chaperone a 2.4 + 0.0 TFAM transcription factor A, mitochondrial 16.3 + 3.7

TIMM23 translocase of inner mitochondrial membrane 23 7.8 + 1.5

TOMM70A translocase of outer mitochondrial membrane 70 6.5 + 1.3

TUBA1C tubulin alpha 6 8.2 + 8.6

TUBB2C tubulin, beta, 2 5.1 + 0.5

TXN thioredoxin 6.9 + 1.6

TXNRD1 thioredoxin reductase 1 6.6 + 0.2

UBA52 ubiquitin and ribosomal protein L40 precursor 10.8 + 2.4

UBE2K ubiquitin-conjugating enzyme E2-25K isoform 1, 6.3 + 0.8

UBE2N ubiquitin-conjugating enzyme E2N 14.9 + 2.3

UBE2V2 ubiquitin-conjugating enzyme E2 variant 2 5.6 + 1.4

VCL vinculin isoform meta-VCL 11.7 + 2.1

VIL2 villin 2 11.1 + 3.1

VIM vimentin 4.9 + 0.1

Degradation rates were obtained by subtracting the dilution rate (oi_dii=0.03±0.004 1/h, from the removal rates. Degradation was dominant for 45 % of the proteins, dilution was dominant for 12 , and the two were comparable for 43 % (017C). Thus, the dependence of protein removal on degradation or dilution varies widely between proteins. We tested whether this correlates with function and localization (0). Proteins localized to the cytoplasm had higher degradation rates than expected by chance (mHG P<10^"4), as did members of the anaphase promoting complex (mHG P<10 ²) and ceU-cycle regulating proteins (mHG P<10^"3). In contrast, proteins of the translation-initiation complex tended to be degraded more slowly (mHG P<10^"2). Thus, proteins with similar function or localization seemed to rely similarly on either dilution or degradation for their removal.

The present inventors used the anti-cancer drug champtothecin (CPT), a topoisomerase-1 poison, and measured the half-life of additional proteins (above those tested in Example 4) for 24 hours after drug addition.

The results are illustrated in Table 10 herein below.

Table 10

77/2 under 77/2 after CPT

normal growth addition

Gene symbol (hours) (hours)

BAG1 5.3 + 1.6 9.5 + 1.5

CKS2 2.6 + 1.6 1.3 + 0.8 DDX5, 8.7 + 1.0 12.9 + 3.5

DYNC1H1 13.8 + 4.3 25.9 + 3.5

EEF2 8.8 + 0.7 7.9 + 1.8

ENOl 8.0 + 3.5 26.3 + 7.3

GAPDH 0.8 + 0.2 1.7 + 0.4

LMNA 12.8 + 2.2 31.0 + 7.0

MYH9 6.4 + 0.6 13.2 + 3.6

PSMB4 7.3 + 0.8 10.9 + 3.6

RPS3A 6.7 + 1.0 3.2 + 0.6

STMN1 3.9 + 0.6 11.3 + 2.0

FSCN1 5.8 + 0.2 13.1 + 4.0

H2AFV 13.9 + 1.8 >48 + NA

K- -1 15.3 + 2.4 . 30.9 + 8.1

NASP 16.3 + 2.3 >48 + NA

RPS3 5.5 + 2.4 4.2 + 1.6

RPS7 4.3 + 1.3 4.4 + 2.1

VCL 11.7 + 2.1 38.5 + 8.0

VIL2 11.1 + 3.1 23.0 + 4.0

VIM 4.9 + 0.1 16.1 + 3.9

CALM2 6.7 + 1.1 11.5 + 2.9

CD44 20.8 + 1.9 16.0 + 5.0

COTL1 8.4 + 1.4 9.5 + 1.0

DDX18 20.3 + 1.4 22.5 + 4.2

ILF2 16.5 + 3.5 34.1 + 8.1

RFC1 9.3 + 1.5 18.4 + 4.6

RPA2 11.2 + 2.0 21.9 + 4.1

RPL22 11.1 + 2.8 14.9 + 2.3

TARS 6.6 + 1.1 11.6 + 1.4

VPS26A 10.6 + 2.3 15.3 + 6.9

RPS6 14.4 + 3.5 21.3 + 7.8

Comparing protein removal with and without the drug revealed a global increase in half-lives: most proteins (22/32) increased their half-life or retained the same half-life (9/32), and only one showed a decrease. Mean half-life doubled from 9.0±4.6h to 18.8±14.8h (paired t-test P<10^"5). This effect also persisted 24-48 hours after drug addition (Figures 18A-B). Notably, the increase in half-lives showed the following pattern: Long-lived proteins became longer-lived in response to the drug, whereas short-lived proteins remained largely unaffected.

The systemic increase in half-lives means that protein removal rates were globally reduced by the drug. One mechanism that could cause this decrease is down- regulation of the degradation machinery (e.g. inhibition of ubiquitin-proteasome mediated proteolysis). However, this would not account for the differential half-life increase of long-lived proteins. On the contrary, long-lived proteins would be expected to be the least affected, because their half -lives are largely determined by dilution due to cell-growth rather than degradation (Figure 19B, Figure 20).

An alternative explanation is that the drug stopped cell growth, thus reducing the dilution rate. Indeed mitosis rate dropped immediately upon drug addition and halted after 5-7h. To see why growth arrest is sufficient to produce the observed half-life effect note that growth arrest eliminates dilution, which differentially affects the protein removal rate a. Since α=<¾_θ5+α_<ΐϋ, proteins with slow degradation have a larger relative reduction in a than proteins with fast degradation (described by Eq. 4-5 in Figure 19C).

This reasoning can quantitatively predict a protein's half-life after growth arrest, *i 2, based on its half-life prior to the arrest, T1/2, and the average cell-cycle duration, Tec (Eq. 5 in Figure 19Q. Using the measured cell-cycle duration, T_∞ = 22.5+/- 2.6 hours and IC2 = 0, because cell-cycle stops after drug addition, generates predictions for the half-lives after drug addition. -Note, that since the model parameters are all measured, no parameter fitting is required. The predictions (blue line in Figure 9B), capture the measured behavior reasonably well (P<10'⁴, Figure 21). Three proteins (CD44, DDX18 and RPS3A) deviated from the general trend with degradation rates increasing in response to the drug, indicating specific degradation regulation (Figure 22).

To further test the generality of this effect, half-life changes were measured in response to four additional stresses: serum-starvation, the transcription inhibitor actinomycin-D, and the anti-cancer drugs paclitaxel and cisplatin, which reduced cell- division rates to 15%, 10%, 0% and 85% of pre-drug rates, respectively. It was found the same effect as in CPT: Long-lived proteins became longer-lived whereas short-lived proteins remained less affected (0 19D-G, Tables 11-14), in a way that was quantitatively predictable using the appropriately reduced dilution rates.

Table 11 illustrates protein half-lives following addition of the drug paclitaxel.

Table 11

Table 12 illustrates protein half-lives during serum starvation.

Table 12

Table 13 illustrates protein half -lives following addition of the drug cisplatin.

Table 13

Gene 77/2 after cisplatin

symbol addition (hours)

BAG1 5.6 + 0.9

CKS2 0.9 + 1.3

DDX5 6.1 + 1.4 H2AFV 20.5 + 1.6

LMNA 8.0 + 2.1

NASP 15.8 + 2.3

RPS7 2.7 + 0.2

VIL2 14.5 + 6.1

Table 14 illustrated protein half-lives after addition transcription inhibitor actinomycin-D

Table 14

One unexpected finding is that the present cells do not appear to compensate for changes in growth-rate by correspondingly altering protein degradation rates. Consequentially, changes in growth-rate directly affect protein half-lives, resulting in corresponding changes in protein levels (023A). Thus, drugs that change growth-rate can cause a global imbalance in proteome levels. This imbalance is expected to be larger the faster the cells grow prior to the drug (0B). Proteome unbalancing may cause lethality, and therefore might be a mechanism for differential killing of rapidly growing cell types, such as tumors, by means of growth-arresting drugs.

The present analysis also highlights the inherent sensitivity of long-lived proteins to fluctuations in cellular growth (023C), suggesting that one way to preserve robust levels is by maintaining proteins short-lived. Furthermore, to preserve stochiometry, it helps to provide proteins in the same complex or system with similar degradation rates, so that fluctuations in dilution would not affect the ratio of their levels.

Elucidating the principles that govern protein removal is important for understanding how cells dynamically control their proteome. This study presents an accurate assay of protein removal, and a principle by which one can understand and predict how different stresses, including chemotherapeutic drugs, affect protein removal in living human cancer cells.

Although the invention has been described in conjunction with specific embodiments thereof, it is evident that many alternatives, modifications and variations will be apparent to those skilled in the art. Accordingly, it is intended to embrace all such alternatives, modifications and variations that fall within the spirit and broad scope of the appended claims.

All publications, patents and patent applications mentioned in this specification are herein incorporated in their entirety by reference into the specification, to the same extent as if each individual publication, patent or patent application was specifically and individually indicated to be incorporated herein by reference. In addition, citation or identification of any reference in this application shall not be construed as an admission that such reference is available as prior art to the present invention. To the extent that section headings are used, they should not be construed as necessarily limiting.

REFERENCES

Futcher, B., et al., A sampling of the yeast proteome. Mol Cell Biol, 1999.

19(11): p. 7357-68.

Gygi, S.P., et al., Quantitative analysis of complex protein mixtures using isotope-coded affinity tags. Nat Biotechnol, 1999. 17(10): p. 994-9.

Gygi, S.P., et al., Correlation between protein and mRNA abundance in yeast. Mol Cell Biol, 1999. 19(3): p. 1720-30.

Ideker, T., et al., Integrated genomic and proteomic analyses of a systematically perturbed metabolic network. Science, 2001. 292(5518): p. 929-34.

Ghaemmaghami, S., et al., Global analysis of protein expression in yeast.

Nature, 2003. 425(6959): p. 737-41.

Lu, P., et al., Absolute protein expression profiling estimates the relative contributions of transcriptional and translational regulation. Nat Biotechnol, 2007. 25(1): p. 117-24.

Kudla, G., et al., Coding-sequence determinants of gene expression in

Escherichia coli. Science, 2009. 324(5924): p. 255-8.

Lian, Z., et al., Genomic and proteomic analysis of the myeloid differentiation program. Blood, 2001. 98(3): p. 513-24.

Chen, G., et al., Discordant protein and mRNA expression in lung

adenocarcinomas. Mol Cell Proteomics, 2002. 1(4): p. 304-13.

Mootha, V.K., et al., Integrated analysis of protein composition, tissue diversity, and gene regulation in mouse mitochondria. Cell, 2003. 115(5): p. 629-40. Tian, Q., et al., Integrated genomic and proteomic analyses of gene expression in Mammalian cells. Mol Cell Proteomics, 2004. 3(10): p. 960-9.

Shankavaram, U.T., et al., Transcript and protein expression profiles of the NCI-60 cancer cell panel: an integromic microarray study. Mol Cancer Ther, 2007. 6(3): p. 820-32.

Yi, Z., et al., Global relationship between the proteome and transcriptome of human skeletal muscle. J Proteome Res, 2008. 7(8): p. 3230-41. Jarvik, J.W., et al., In vivo functional proteomics: mammalian genome annotation using CD-tagging. Biotechniques, 2002. 33(4): p. 852-4, 856, 858-60 passim.

Sigal, A., et al., Dynamic proteomics in individual human cells uncovers widespread cell-cycle dependence of nuclear proteins. Nat Methods, 2006. 3(7): p. 525-31.

Sigal, A., et al., Variability and memory of protein levels in human cells. Nature, 2006. 444(7119): p. 643-6.

Sigal, A., et al., Generation of a fluorescently labeled endogenous protein library in living human cells. Nat Protoc, 2007. 2(6): p. 1515-27.

Cohen, A. A., et al., Dynamic proteomics of individual cancer cells in response to a drug. Science, 2008. 322(5907): p. 1511-6.

Pratt, J.M., et al., Dynamics of protein turnover, a missing dimension in proteomics. Mol Cell Proteomics, 2002. 1(8): p. 579-91.

Yen, H.C., et al., Global protein stability profiling in mammalian cells. Science, 2008. 322(5903): p. 918-23.

Pommier, Y., Topoisomerase I inhibitors: camptothecins and beyond. Nat Rev Cancer, 2006. 6(10): p. 789-802.

Uri, A., An Introduction to Systems Biology: Design Principles of Biological Circuits 2007: CRC press.

Belle, A., et al., Quantification of protein half-lives in the budding yeast proteome. Proc Natl Acad Sci U S A, 2006. 103(35): p. 13004-9.

Surrey, T., et al., Chromophore-assisted light inactivation and self-organization of microtubules and motors. Proc Natl Acad Sci U S A, 1998. 95(8): p. 4293-8. Tour, O., et al., Genetically targeted chromophore-assisted light inactivation. Nat Biotechnol, 2003. 21(12): p. 1505-8.

Hampton, R.Y., et al., In vivo examination of membrane protein localization and degradation with green fluorescent protein. Proc Natl Acad Sci U S A, 1996. 93(2): p. 828-33.

Morin, X., et al., A protein trap strategy to detect GFP-tagged proteins expressed from their endogenous loci in Drosophila. Proc Natl Acad Sci U S A, 2001. 98(26): p. 15050-5. Clyne, P.J., et al., Green fluorescent protein tagging Drosophila proteins at their native genomic loci with small P elements. Genetics, 2003. 165(3): p. 1433-41.

Huh, W.K., et al., Global analysis of protein localization in budding yeast. Nature, 2003. 425(6959): p. 686-91.

Jarvik, J.W., et al., CD-tagging: a new approach to gene and protein discovery and analysis. Biotechniques, 1996. 20(5): p. 896-904.

Jarvik, J.W. and C.A. Telmer, Epitope tagging. Annu Rev Genet, 1998. 32: p. 601-18.

Fisher, R.A., The Use of Multiple Measurements in Taxonomic Problems.

Annals of Eugenics, 1936. 7: p. 179-188.

Adams, R. and L. Bischof, Seeded Region Growing. IEEE Trans Pattern Anal Mach Intell, 1994. 16(6): p. 641-647.

Eden, E., et al., An automated method for analysis of flow characteristics of circulating particles from in vivo video microscopy. IEEE Trans Med Imaging, 2005. 24(8): p. 1011-24.

Downey, T., Analysis of a Multifactor Microarray Study Using Partek Genomics Solution Methods in Enzymology, 2006. 411: p. 256-270.

Ashburner, M., et al., Gene ontology: tool for the unification of biology. The Gene Ontology Consortium. Nat Genet, 2000. 25(1): p. 25-9.

Eden, E., et al., GOrilla: a tool for discovery and visualization of enriched GO terms in ranked gene lists. BMC Bioinformatics, 2009. 10: p. 48.

Eden, E., et al., Discovering motifs in ranked lists of DNA sequences. PLoS Comput Biol, 2007. 3(3): p. e39.

Claims

WHAT IS CLAIMED IS:

1. A method of determining a degradation rate of a fluorescently labeled polypeptide of interest in a population of cells, the method comprising:

(a) obtaining two identical populations of cells;

(b) determining fluorescence levels of the fluorescently labeled polypeptide of interest in a first of said at least two populations of cells at at least two different points in time;

(c) at least partially reducing a fluorescence of the fluorescently labeled polypeptide of interest in a second of said at least two populations of cells;

(d) determining fluorescence levels of said reduced fluorescently labeled polypeptide of interest in said second of said at least two populations of cells at said at least two different points in time;

(e) comparing fluorescence levels in said first and said second populations; and

(f) based on said comparing, determining a degradation rate of the fluorescently labeled polypeptide of interest.

2. The method of claim 1, wherein step (f) is effected by calculating the ratio of the (natural log (In) of the difference in fluorescence of said polypeptide of interest in said first population of cells and fluorescence of said polypeptide of interest in said second population of cells at a first point in time) : (In of the difference in fluorescence of said polypeptide of interest in said first population of cells and fluorescence of said polypeptide of interest in said second population of cells at a second point in time) divided by the difference in time between said two points in time.

3. The method of claim 1, wherein step (f) is effected by determining a slope of a graph, wherein a y axis of said graph indicates the In (P(t) -Pv(t)) and an x axis of said graph indicates time, wherein P(t) is the amount of fluorescence of said polypeptide of interest at point t in time in said first population of cells and (Pv(t) is the amount of fluorescence of said polypeptide of interest at point ί in time in said second population of cells.

4. The method of claim 1, wherein step (c) is effected with a mercury fluorescent lamp.

5. The method of claim 1, wherein step (c) is effected for 1-8 minutes.

6. The method of claim 1, wherein said population of cells comprises an additional fluorescently labeled polypeptide, said additional fluorescently labeled polypeptide having a higher nuclear: cytoplasm expression ratio and said additional fluorescently labeled polypeptide being distinguishable from said fluorescently labeled polypeptide of interest.

7. The method of claim 1, wherein the polypeptide of interest is transcribed from its native location in said population of cells.

8. A method of analyzing the effect of an agent on a degradation rate of a polypeptide of interest in a population of cells, the method comprising:

(a) contacting a population of cells with the agent;

(b) determining the degradation rate of said fluorescently labeled polypeptide of interest in said population of cells according to the method of claim 1; and

9. A system for determining a degradation rate of a fluorescently labeled polypeptide of interest in a cell comprising a processing unit, said processing unit executing a software application configured for converting change in fluorescence levels over a period of time to a degradation rate.

10. The system of claim 9, said degradation rate is determined according to the method of any of the claims 1-7.

11. The system of claim 9, further comprising an imaging system.

12. The system of claim 11, wherein said imaging system comprises an image capture apparatus.

13. The system of claim 12, wherein said image capture apparatus is capable of capturing an image of said fluorescently labeled polypeptide over said period of time.

14. The system of claim 11, wherein said imaging system further comprises a data processor for calculating a fluorescence level for each picture element of said image.

15. The system of claim 12, wherein said image capture apparatus is a fluorescent image capture apparatus.