Identification and characterization of post-translational modifications on histones: an on-line top-down mass spectrometry workflow for analysis using ProSight PTM

(USC Thesis Other)

Identification and characterization of post-translational modifications on histones: an on-line top-down mass spectrometry workflow for analysis using ProSight PTM

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content IDENTIFICATION AND CHARACTERIZATION OF POST-TRANSLATIONAL MODIFICATIONS ON HISTONES: AN ON-LINE TOP-DOWN MASS SPECTROMETRY WORKFLOW FOR ANALYSIS USING PROSIGHT PTM by Heather Noelle Reilly-Rhoten A Thesis Presented to the FACULTY OF THE USC GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree MASTER OF SCIENCE (BIOCHEMISTRY AND MOLECULAR BIOLOGY) May 2012 Copyright 2012 Heather Noelle Reilly-Rhoten ii Table of Contents List of Tables iii List of Figures iv Abbreviations vi Abstract vii Introduction Mass Spectrometry 1 Proteomics and Mass Spectrometry 6 Histones 8 Chapter One: Workflow Summary 13 Chapter Two: Histone Identifications 19 Identification of Histone 2A 20 Identification of Histone 2B 24 Identification of Histone 3 28 Identification of Histone 4 32 Chapter Three: Testing the Workflow with Purified Histone H2A/B 36 Chapter Four: Analyzing a Mixture of Histones Analysis of a Mixture of Histones-Experiment 1 38 Analysis of a Mixture of Histones-Experiment 2 40 Analysis of a Mixture of Histones-Experiment 3 42 Analysis of a Mixture of Histones-Experiment 4 45 Analysis of a Mixture of Histones-Experiment 5 48 Analysis of a Mixture of Histones-Experiment 6 50 Summary and Conclusions 52 References 54 iii List of Tables Table 1: Histone Matches and Modifications for Purified H2A/B 37 Table 2: Histone Matches and Modifications for Experiment 1 39 Table 3: Histone Matches and Modifications for Experiment 3 43 Table 4: Histone Matches and Modifications for Experiment 4 46 Table 5: Histone Matches and Modifications for Experiment 5 48 Table 6: Histone Matches and Modifications for Experiment 6 51 Table 7: Summary of Histones Identified in Experiments 52 Table 8: Summary of Post-Translational Modifications Characterized on the Histone Matches 52 iv List of Figures Figure 1: Types of Fragment Ions 4 Figure 2: Organization of DNA within the Nucleus 8 Figure 3: Determination of a Peptide Sequence by Tandem Mass Spectrometry 10 Figure 4: Spectrographs of a MS and MS 2 Experiment 14 Figure 5: Deconvolution of MS Data 15 Figure 6: Search Window of ProSight PTM 16 Figure 7: Search Results List from ProSight PTM 17 Figure 8: Spectrographs of MS and MS 2 for H2A 20 Figure 9: Deconvolution of MS Data for H2A 21 Figure 10: Search Results from ProSight PTM for H2A 22 Figure 11: Fragment Mapping and Modifications for H2A 23 Figure 12: Spectrographs of MS and MS 2 for H2B 24 Figure 13: Deconvolution of MS Data for H2B 25 Figure 14: Search Results from ProSight PTM for H2B 26 Figure 15: Fragment Mapping and Modifications for H2B 27 Figure 16: Spectrographs of MS and MS 2 for H3 28 Figure 17: Deconvolution of MS Data for H3 29 Figure 18: Search Results from ProSight PTM for H3 30 Figure 19: Fragment Mapping and Modifications for H3 31 Figure 20: Spectrographs of MS and MS 2 for H4 32 Figure 21: Deconvolution of MS Data for H4 33 v Figure 22: Search Results from ProSight PTM for H4 34 Figure 23: Fragment Mapping and Modifications for H4 35 vi Abbreviations ATP= Adenosine Triphosphate CID= Collision-Induced Dissociation DNA= Deoxyribonucleic Acid ESI= Electrospray Ionization ECD= Electron-Capture Dissociation ETD= Electron-Transfer Dissociation FTMS= Fourier Transform Ion Cyclotron Resonance Mass Spectrometry HCD= Higher-Energy C-Trap Dissociation MALDI= Matrix-Assisted Laser Desorption/Ionization MS= Mass Spectrometry m/z= Mass to Charge Ratio NSI= Nanospray Ionization RT= Retention Time TOF= Time-of-Flight Spectrometry vii Abstract The importance of histones for DNA management cannot be understated. From rearrangement to regulation of expression, histones act as the principle effectors of DNA availability and storage. A majority of diseases involve errors in gene expression; frequently chemical changes to the DNA and histones comprising the epigenome are to blame. Mass spectrometry has recently exploded as a prime method for studying histones. Advancements in the resolution and sensitivity of mass spectrometry machinery have allowed histone investigation to expand beyond traditional protocols. Namely, the ability to study intact histones without the need for enzymatic digestion beforehand has allowed more information to be gained about histones, which was previously unobtainable. An on-line intact histone workflow was developed to establish a protocol for identifying different histone types in a sample. Characterization of chemical modifications on the histones was accomplished using an online version of the ProSight PTM software. From this method, the four histones of the nucleosome octamer were identified and several post-translational modifications were detected. Expanded application of the developed workflow could be used to better understand known modifications and detect novel modifications. Additional adaptations of this analytical procedure could be used to test the epigenetic signals between different physiological conditions and find indicators of disease states. The initial success of the developed methodology for identifying modifications to histones provides the foundation for a rapid expansion of knowledge about the human epigenome. 1 Introduction Mass Spectrometry Mass spectrometry (MS) comprises of ionizing samples, selecting for certain ions, and detecting the mass to charge ratio (m/z) of the ions. Possible information that can be gained about a sample from analysis by mass spectrometry includes sequence, mass, structure, modifications, quantification, and stability, to name a few. Off-line analysis involves fractionation of a sample by various chromatography methods then introduction of the fractions separately into the mass spectrometer. On-line analysis of a sample refers to a sample that is flowed through a chromatography column attached to the mass spectrometer in order to fractionate the components of the sample before they are directly ionized and sent into the mass spectrometer for detection. There are two main methods for ionizing samples for induction into the spectrometer: matrix-assisted laser desorption/ionization (MALDI) and electrospray ionization (ESI). With the MALDI set-up, a matrix substance is mixed with the sample in excess. The matrix absorbs the ultraviolet from the laser, providing the necessary energy to vaporize and ionize a bit of the sample, which is then directed into the MS machine for detection. Conversely, ESI operates by sending the sample, mixed in an appropriate solvent, through a small, charged spray tip. As the sample/solvent mixture leaves the tip, it bursts into tiny charged droplets. The solvent evaporates as the sample travels to the MS machine for detection. An unsustainable amount of Coulombic force remains on the droplets, causing them to burst into individually charged ions which then enter the 2 machine. There are numerous others types of ionization; the properties of the analyte determine which types are appropriate for use. Next, the ions need to be sorted and selected for detection. This can be effected via various different types of traps. Quadrupole linear ion traps arrest ions as they flow into a tunnel-like structure created by four charged rods. Constant voltage changes of the rods keep the charged molecules confined to the tunnel, allowing the experimenter to choose which ions to send to the detector. The detector is usually an electron multiplier which releases electrons in response to a charged particle hitting its surface, starting a cascade of electrons that are interpreted by a computer into an m/z spectrum. Alternately, an Orbitrap holds ions by oscillating them around a centralized rod using the interactions of the ions with the charged electrode and centripetal force. Voltage changes inside of the Orbitrap allow the experimenter to select specific ions to be detected. Ions selected for detection are oscillated near metal plates that record the current produced by the ions, which are then Fourier transformed and interpreted by a computer into an m/z spectrum (Hu et al. 2005). Another type of selection process frequently used for mass spectrometry is called time-of-flight (TOF). This method uniformly accelerates ions of the same charge across a known distance to a detector. Differences in the masses of the ions produce a natural separation among the ions, with large, heavy ions taking the longest time to reach the detector. From the travel time to the detector, the exact m/z of each ion is determined. The experimenter uses a velocity filter to block unwanted ions in order to narrow the pool of ions sent for detection. A micro-channel plate detector with numerous small electron 3 multiplier cells is used to detect the ions’ charges to be converted by a computer into an m/z spectrum. While these selection and detection set-ups are the prominent systems employed for mass spectrometry, there are various other types of traps and analyzers that are used in the discipline. In order to investigate an analyte further, the ions can be subjected to various fragmentation methods. The m/z values of ion fragments can be used to identify the molecules and sometimes determine the sequence or structure of a molecule. Portions of the molecule are neutralized or cleaved off by various methods, leaving behind charged fragments of the original molecule that are sent to the MS detector. With MALDI, fragmentation is achieved by increasing the laser intensity or changing the matrix to be more adaptive for producing fragments. Fragmentation occurs in ion traps and Orbitraps by methods such as collision-induced dissociation (CID), higher-energy C-trap dissociation (HCD), electron-capture dissociation (ECD), and electron-transfer dissociation (ETD). All the systems inside the mass spectrometer use third-party gases to break bonds or to charge molecules and neutralize portions of the analyte by transferring charges. For example, ETD uses the chemical fluoranthene to transfer negative charges to the analyte ions. Solid fluoranthene is heated in the ETD module until it is vaporized as radical anions (Chi et al. 2007). The fluoranthene anions travel to the trap storing the analyte ions, where they react with the analyte, causing fragmentation and a transfer of negative charge. Neutralized fluoranthene molecules are flushed from the system by the vacuum in the machine. This method is particularly beneficial for investigating analytes 4 with numerous positive charges, as reaction with multiple fluoranthene anions can lower the charge of the fragments into a detectable m/z range. The initial ions undergoing fragmentation (precursor ions) end up as separate charged pieces following fragmentation (product ions). For proteins and peptides, these different product ions depend on which bonds are randomly broken along the backbone and what end of the molecule retains the charge. Charges remaining on the N-terminus of the peptide fragment are x-, y-, or z-type ions and charges remaining on the C-terminus are a-, b-, or c-type ions. Along the peptide backbone, x- and a-type ions are broken at the C Į n and C=O bond, b- and y-type ions are broken at the C=O and N-H bond, c- and z- type ions are broken at the N-H and C Į n+1 bond (Figure 1 (Roesporff and Fohlman 1984) ). Figure 1: Types of Fragment Ions When first-level precursor ions produced by the mass spectrometer are selected and detected, the experiment is called MS or MS 1 . The second-level product ions created by fragmentation of the precursor ions are also selected for and detected by the mass spectrometer (MS 2 also called tandem mass spectrometry). Subsequent fragmentation of 5 the product ions leads to third-level ions which can also be analyzed by the mass spectrometer (MS 3 ). Successive fragmentations lead to optional multi-level analysis limited by the molecule being analyzed and the capacity of the particular mass spectrometer machine (MS n ). 6 Proteomics and Mass Spectrometry The recent rapid expansion of the field of protein mass spectrometry has been largely due to improvements in the structure and functionality of advanced mass spectrometry hybrid systems. One major factor contributing to the rise of mass spectrometry for proteomic studies is the improvement of the mass spectrometer machine itself and its coupling to related instrumentation. The enhanced resolution, speed, and detection abilities of Thermo Fisher Scientific’s Orbitrap technology give researchers the finesse to measure low abundance and high molecular weight proteins. Coupling Orbitrap sensitivity to the utility of ETD gives even more flexibility for studying proteins with high charge states and modifications that are unobtainable using other methods like CID. CID produces mainly b- and y-type ions and causes loss of modifications due to internal vibrational energy transfer (Hoffert and Knepper 2008). With ETD, the protein is fragmented along the protein backbone forming c- and z-type ions, leaving most modifications intact (Syka et al. 2004). Analyses of proteins with hybrid machine systems yield more information than any method used separately could provide, leading to a better understanding of the protein under investigation. Along with improvements to the machinery, there have been correlated improvements in MS related software. Changes to spectra analysis algorithms have been constantly upgraded to accommodate the complexity of the data produced by advanced hybrid MS systems. Also, specialized software has been developed to address specific biological questions and topics. One such software, ProSight PTM, created by the Kelleher lab formerly of the University of Illinois at Urbana-Champaign, surfaced as a 7 promising proteomics tool for comparison of tandem MS data to proteomic sequence databases. In particular, ProSight PTM was developed for “identification and characterization of intact proteins and their post-translational modifications” (Kelleher Research Group 2011). Because it utilizes the extensive UniProt proteomic databases and provides a create-your-own database option, ProSight PTM offers search flexibility and specificity for identifying matching sequences to mass spectrometry data. ProSight PTM also supports the use of a variety of fragmentation methods for producing MS n fragments to analyze including ETD, ECD, and HCD. An important search feature of ProSight PTM allows protein sequences to be searched for numerous known post-translational modifications. If a modification is unknown or poorly characterized, a delta mode search can be performed to find recurring mass shifts in the MS n fragments. After analysis, ProSight PTM provides maps of modifications along the sequence of the protein. Overall, ProSight PTM serves as a potent tool for identifying intact proteins and characterizing their modifications. Improvements to the machines and software have expanded the applications of mass spectrometry across many new and developing topics of scientific interest. For example, in the emerging area of proteogenomics, mass spectrometry has been playing a key role in bridging the gap between the proteome and genome for gene annotation (Gupta et al. 2008). 8 Histones Histones perform an integral role as regulators in the cell by modulating DNA expression, activity, and arrangement and are hence the subject of increasing amounts of investigation. Histones serve as a structural matrix for DNA condensation by providing an octamer of proteins the DNA wraps around to form structures known as nucleosomes. Nucleosomes pack together to form higher order arrangements called chromatin. Strands of chromatin wrap around each other to form rope-like fibers which bundle tightly together during mitosis to create the characteristic X-shaped chromosome. Condensation of DNA into higher order conformations allows a large amount of DNA to package into a much smaller nucleus (Figure 2 (Epitron 2012) ). Each histone octamer consists of two subunits of the histone proteins H2A, H2B, H3, and H4. Figure 2: Organization of DNA within the Nucleus Additionally, there exist variants of H2A and H3 which replace their counterparts in the nucleosome structure to promote certain processes such as DNA repair, activation, 9 repression, and stability. Replacement by histone variants acts as one of the three main ways of regulating DNA by histones. A second method of DNA regulation through histones uses ATP to reshuffle the nucleosome structures along the DNA. Changes to nucleosome locations alter the available stretches of DNA open to the cellular environment and accessible for transcription or modification. The term euchromatin refers to chromatin arranged in a loose conformation with stretches of accessible DNA able to be actively transcribed. Conversely, heterochromatin is compacted chromatin with DNA inaccessible for transcription. Enzymes use ATP to rearrange euchromatin and heterochromatin regions of DNA in order to regulate gene expression in the cell. Finally, a major mechanism for regulating gene activity entails modifying key residues on the histone proteins to promote gene activation or repression. Most of the known modifications occur on the tail-end section of the histone proteins that protrude from the nucleosome structure. By extending beyond the wrapped DNA, the histone tails are within reach of the enzymes that alter them. Some enzymes develop mechanisms for entering the core of the nucleosome structure to modify internal residues on the histones, but the majority act on the histone tails. Common modifications to histone residues include acetylation, methylation, phosphorylation, and ubiquitination. Residues can accommodate single or multiple modifications at the same site. Each modification affects the activation or repression of the genes located on the DNA wrapped on that histone octamer by influencing histone-DNA and histone-histone interactions. 10 Some of the modifications to histone residues have been found to be transmitted during replication and are inherited by the daughter cells. This indicates the importance of the modifications to the functionality of the cell. Virtually all cellular processes rely on the activation and repression of genes to function. Specifically, regulation by histones often critically affects cellular differentiation and maintenance. The management of histones encompasses an important area of research as numerous diseases are caused by or can be treated by mechanisms altering the histone activity at specific gene sites. Two basic approaches are commonly taken when investigating protein or peptide samples such as histones using mass spectrometry. A bottom-up approach involves digesting the protein or peptide with a restriction enzyme, usually trypsin, to introduce smaller fragments to the MS machine. The fragments are detected and the sequence of the original molecule can be found using the mass differences between overlapping fragments to calculate the amino acid sequence (Figure 3 (Addona and Clauser 2002) ). Figure 3: Determination of a Peptide Sequence by Tandem Mass Spectrometry A bottom-up method is useful for large-scale protein studies and complex mixtures of proteins since the separation of peptides is better than that of intact proteins. 11 However, protein modifications can incur damage through the digestion process, making a bottom-up method not ideal for studying modified proteins. Alternatively, researchers can use a top-down method for protein analysis in which the protein is injected into the machine intact, without digestion. This method requires a mass spectrometer with advanced resolving power and sensitivity, but is gentler on modifications and generates better sequence coverage (Yates et al. 2009). Many studies investigating histones in the past have applied a bottom-up approach, but as a result, lost valuable information about labile modifications. Enhancement of the top-down approach by utilization of ETD for fragmentation preserves the integrity of histone modifications, allowing them to be detected during analysis. A major advantage of using of using mass spectrometry to study histones is the amount of information acquired from a single experiment. Other techniques for studying proteins have substantial limitations and can only uncover a few details about the analyte. For example, 2D electrophoresis is a time consuming method with low reproducibility and does not separate large or hydrophobic proteins well (Rabilloud 2009). In Western blotting, there is poor transference of intact proteins with high molecular weight to the membrane resulting in reduced detection abilities. With sequencing by Edman degradation, proteins must be broken down into short sequence fragments in order to obtain a reliable sequence. Protein sequence prediction from DNA or mRNA has difficulty predicting alternative splicing and does not account for any post-translational changes to the proteins. Mass spectrometry allows the flexibility to study histones intact, yield sequence information, and detect modifications to the histone sequences. 12 By running the histone sample on-line, the mass spectrometer portion of the experiment can be fully automated providing a reduction in the time from sample injection to the interpretation of meaningful results. In addition, there is less handling of the sample, leading to reduced contamination and cleaner results. The purpose of the following investigation was to develop a cohesive method to run intact histones through a top-down method, identify the histone species present by tandem mass spectrometry, and characterize the post-translational modifications on the histones using ProSight PTM software. This methodology provides a quick turnaround for large-scale analysis of histones and has the potential to generate significant contributions to the field of epigenetics. 13 Chapter One: Workflow Summary For analysis of histones using a top-down approach, a method was established to identify histones types present in a sample and characterize the modifications found along the histone sequences. First, 10 μL of histone solution was direct-injected into a Thermo Scientific LTQ Orbitrap XL ETD hybrid mass spectrometer. The sample flowed through either a self-made C18-packed reverse-phase silica column (10 cm of resin, 400x0.1 mm) or a Jupiter C4-packed (5 ȝ m, 300Å, 250x0.3 mm, Phenomenex, Torrance, CA) silica column to a nanospray (NSI) source with either a C8-packed tip (filled 2 cm), C18- packed tip (filled 2 cm), or an empty tip. Once ionized, the histones were sent to the FTMS to be detected in the positive mode over a range of 375- 1600 m/z. The aqueous phase (A) used for the gradient was composed of HPLC-quality water with 0.1% formic acid and the organic phase (B) was composed of HPLC-quality acetonitrile with 0.1% formic acid. Several parameters of the mass spectrometer were changed to accommodate intact protein analysis. Concurrent to the MS analysis of the precursor ions, the top five most intense precursor ions of each scan were subjected to ETD fragmentation in the FTMS. After fragmentation, those precursor ions were prohibited from being fragmented again for at least 20 s thereby allowing for fragmentation of other precursor values. After the completion of the run, the MS and MS 2 spectra were viewed in Thermo Scientific’s Qual Browser (Figure 4). Probable histone elution envelopes during the run were localized and spectra from specific retention times (RT) during elution were chosen for analysis. 14 Similarly, ETD-induced fragment spectra occurring during histone elution were chosen from the MS 2 spectrograph for alignment with the precursors. Figure 4: Spectrographs of a MS and MS 2 Experiment On the spectrographs generated by the mass spectrometer, the top graph is a function of the relative abundance of the eluting materials over the course of the run. The x-axis is the time/retention duration of the material. The second red graph is the same, except that it illustrates the ETD fragment ion intensities throughout the run. Next is a graph of a MS spectrum at a particular RT during the run. Details are given about the MS ions detected during the scan such as m/z ratio, intensity, and resolution. Regular high multiple charges within the 500-1100 m/z range were used as indicative factors of possible histone elution envelopes corresponding to the total molecular weight of the histones. The final graph demonstrates similar characteristics, but for MS 2 ions instead of MS ions. Details specific to the product ions include the strength of ETD used for 15 fragmentation and the precursor ion that was fragmented. The next step of analysis was deconvolution of the spectra data to obtain the masses of the ions. For identification of the analyte, the exact masses of the precursors and fragments need to be found to use for comparison to known molecules in search databases such as the UniProt Knowledgebase, NCBI RefSeq, or Protein Data Bank. In order to use the values from the spectra for identification of the histones, the m/z values must be converted into neutral molecular masses. This process is called deconvolution. The Thermo Scientific support software package for the mass spectrometer includes the program Xtract, which is used for deconvoluting MS spectra. Deconvolution is based on the following equation: neutral mass = (m/z) (z)- (z) (mass of H + ). The individual scans chosen from the MS and MS 2 spectra for histone identification were run through the Xtract program to obtain deconvoluted neutral masses (Figure 5). Figure 5: Deconvolution of MS Data 16 The top graph depicts the neutral mass values of the intact histones detected in the sample. Displayed in the second graph is a list of the deconvoluted masses ranked by their intensity. Next, up to 30 of the most intense deconvoluted masses from each precursor peak and each fragment spectra were uploaded into the online ProSight PTM software. Using the Highly Annotated Absolute Mass Search feature of ProSight PTM, the MS and MS 2 masses were run against a H. sapiens database obtained from the UniProt Consortium to search for sequence matches to the mass spectrometer data (Figure 6). Figure 6: Search Window of ProSight PTM 17 On the search page, the first step was to choose the appropriate MS data file and MS 2 data file to compare. Next, the search parameters were narrowed to return meaningful results. Third, the database to search for matching sequences was chosen. Finally, all the available post-translational modifications were selected to be checked for during the program analysis. The search results generated were displayed as a list of reports for each precursor mass which could then be selected for information about sequence matches and fragment comparisons (Figure 7). Figure 7: Search Results List from ProSight PTM Results reported by ProSight PTM included possible sequence matches, number of fragments matching each sequence, alignment of fragments along the match sequence, probability of match accuracy, and modifications found along the match sequence. Modifications were calculated by the shift in fragment masses due to addition of the known masses of specific modifications. By following this workflow, the histone 18 identities were determined and modifications to the histones were characterized by the ProSight PTM software. 19 Chapter Two: Histone Identifications The following four sections (Figures 8-23) give pictorial evidence for the successful identification of the four major histone types (H2A, H2B, H3, and H4) using the workflow discussed as well as detecting modifications along their sequences. The first graph is the spectrograph of the run for the MS and MS 2 . Second is the graph after deconvolution with a corresponding list of neutral masses. Results from the ProSight PTM search are displayed in the third graph. Last is an example of the fragment alignments to the histone matches with modified amino acids in red. 20 Identification of Histone 2A Figure 8: Spectrographs of MS and MS 2 for H2A 21 Figure 9: Deconvolution of MS Data for H2A 22 Figure 10: Search Results from ProSight PTM for H2A 23 Figure 11: Fragment Mapping and Modifications for H2A 24 Identification of Histone 2B Figure 12: Spectrographs of MS and MS 2 for H2B 25 Figure 13: Deconvolution of MS Data for H2B 26 Figure 14: Search Results from ProSight PTM for H2B 27 Figure 15: Fragment Mapping and Modifications for H2B 28 Identification of Histone 3 Figure 16: Spectrographs of MS and MS 2 for H3 29 Figure 17: Deconvolution of MS Data for H3 30 Figure 18: Search Results from ProSight PTM for H3 31 Figure 19: Fragment Mapping and Modifications for H3 32 Identification of Histone 4 Figure 20: Spectrographs of MS and MS 2 for H4 33 Figure 21: Deconvolution of MS Data for H4 34 Figure 22: Search Results from ProSight PTM for H4 35 Figure 23: Fragment Mapping and Modifications for H4 36 Chapter Three: Testing the Workflow with Purified Histone H2A/B To start, an experiment was run to see if a known purified histone could be analyzed using the described workflow and correctly identified and characterized by the ProSight PTM software. Readily available purified human H2A/B from Dr. Hsieh’s lab at University of Southern California was obtained and run via direct injection through a NSI source at a flow rate of 500 nL/min into the mass spectrometer. Once stable elution of a histone-like envelope occurred, the spectrograph was recorded. One major criterion for determination of possible H2A/B histone elution was the presence of high multiply charged m/z values between +6 and +21. Another attribute used in identification of H2A/B histone elution were m/z values around the range of 1000-1900. A peak from the elution envelope at m/z 1148 was isolated for fragmentation to be used for identification of the histone species. The 1148 peak was subjected to ETD and the spectrum of the fragments was recorded. Exact masses of the 1148 precursor and its product ions were found by conversion of the data using Xtract. The deconvolution was run on m/z values between 1145.5 and 1150.0 for the MS mass determination. This resulted in three masses of 13736.51384, 13741.44930, and 13749.45174 Da being ascertained from the 1148 peak. The top 50 most intense peaks from the ETD fragmentation of the 1148 peak were deconvoluted between the m/z values of 700 and 1700. Following deconvolution of the MS and MS 2 data, the calculated neutral masses were exported to the ProSight PTM software online. The database findings from the software for the three precursor masses are summarized in Table 1 below. The protein 37 eluting at m/z 1148 was correctly identified by ProSight PTM as H2B from the sequence determined by the precursor and product ions. From the ProSight results, no modifications were found on these H2B sequences. Table 1: Histone Matches and Modifications for Purified H2A/B Mass (Da) Histone Match Length (a.a.) Mass Difference (Da) Matched Ions Probability Score Modifications 13749.5 Human H2B P62807, P02278, Q93078, Q93 080, Histone H2B.a/g /h/k/l H2B.1 A H2B/a H2B/g H2B/h H2B/k H 2B/l 125 -17.065 7 6.96521e -10 None 13741.4 Human H2B P62807, P02278, Q93078, Q93080, Histone H2B.a/g /h/k/l H2B.1 A H2B/a H2B/g H2B/h H2B/k 125 -25.0675 8 1.47496 e -11 None 13736.5 Human H2B.q Q16778, Histone H2B.q H2B/q 125 -44.0186 9 2.77767 e -13 None This experiment established that a top-down method coupled with ETD fragmentation could produce a correct identification of histones using ProSight PTM software with reasonable accuracy. The lack of modifications found could simply be due to the fact that there were not any on the m/z 1148 ions. Conversely, the lack of modifications could have resulted from damage to the modifications during processing or fragmentation. In spite of this uncertainty and in light of the correct identification of H2B, it was decided to continue on with the next level of experiments and determine if a mixture of all the histone types could be resolved and identified by ProSight PTM and checked for modifications. 38 Chapter Four: Analyzing a Mixture of Histones Analysis of a Mixture of Histones- Experiment 1 Since the established top-down parameters were able to identify a known histone, the next test was to inject a mixture of all types of histones to see if multiple types could be discerned and identified by ProSight PTM. A mixture of intact histones was obtained from Dr. Hsieh’s lab at University of Southern California which had been isolated from human cells via fractionation. Direct injection of 10 ȝ L of the histone mixture was introduced into the mass spectrometer. The sample flowed through a self-made silica column with 10 cm of C18 resin at 500 nL/min to separate the different histone types so each would enter the mass spectrometer at different times. After following the previously described method of analysis and fragmentation in the mass spectrometer, the spectrograph was searched for possible histone elution envelopes. Precursor spectra at RT 54.74, 55.08, and 57.40 were chosen for deconvolution and run through the Xtract software. Corresponding MS 2 spectra were chosen and deconvoluted as well. Two types of histones, H3 and H4, were identified by ProSight PTM from the chosen spectra. Additionally, several types of modifications were detected along the sequences of the identified histones (Table 2). These results were positive for indicating the potential to gain information by the workflow developed. 39 Table 2: Histone Matches and Modifications for Experiment 1 Mass (Da) Histone Match Length (a.a.) Mass Difference (Da) Matched Ions Probability Score Modifications 15372.1 Human H3.1 P16106, P02295, P02296, Histone H3.1 H3/a H3/c H3/d H3/f H3/h H3/i H3/j H3/k 135 +66.6091 9 5.11422e -12 N6-methyl-L- lysine 11360.5 Human H4 P62805, P02304, P02305, Histone H4 102 -22.9418 4 9.3948 e -05 N-acetyl-L-serine N6,N6-dimethyl- L-lysine N2-acetyl-L-lysine The discovery of modifications to the histones by the ProSight PTM software were evidence that the workflow could be used to identify and characterize specific intact histones from a mixture. However, there were only two types of histones found in this experiment. One possible explanation for this occurrence may have been that as the experiment was running, the back pressure in the column was high, indicating there may have been a blockage in the column or tip from the bulky intact histones. This would lead to a decrease in the amount of histones eluting from the column and entering the mass spectrometer for detection. Also, the histones may not have separated enough in the column and eluted at roughly the same time, causing the blockage. A few changes to the system set-up were made to increase the number of histones detected by causing better and smoother histone elution. 40 Analysis of a Mixture of Histones- Experiment 2 For the second run of the histone mixture, a few parameters were adjusted to reduce the back pressure of the system while also increasing the number of histone types detected. One of the changes was dropping the flow rate to 300 nL/min to relieve the back pressure in the system. Doing this would also help the histone separation, which would reduce the clogging of the column and tip. One potential negative side effect of dropping the flow rate is a drop in signal intensity because the elution of each histone is broader and more spatially separated. In order to combat the potential loss of signal, the silica tip was packed with 2cm of C8-resin. With a packed tip, each broad histone segment eluting from the column would be reconcentrated just prior to entering the mass spectrometer for detection, thereby maximizing the amount of histones detected in each scan. Inclusion of a packed tip does increase the back pressure of the system, although the decrease in the flow rate would offset the difference. Another change to the experimental conditions was the addition of 1% (v/v) acetic acid to the histone sample before injection. This was also done for all subsequent experiments. The reasoning for adding acid to the sample was to drop the pH of the sample solution to cause the sample to take on a positive charge. A positive charge would facilitate the flow of the histones through the column because there would be fewer electrostatic interactions between the sample, the walls of the column, and the resin in the column. This experiment ended unsuccessfully. The back pressure in the system continued to rise throughout the run until it exceeded the safe running conditions of the machinery 41 and caused the run to abort. Histone elution had not begun before the run was terminated. Addition of the packed tip may have contributed more pressure than the drop in flow rate could compensate for along with the size of the histones. Alternatively, the tube lens value of the MS machine may have been too low for highly charged molecules to pass into the machine for analysis. The tube lens focuses and accelerates the stream of ions entering into the mass spectrometer to direct the ions toward the trap. Without a high enough voltage setting, the tube lens would not apply the voltage necessary to guide multiply charged molecules into the MS trap. Overall, a better balance between the separation of the histones to prevent blockage and an increase in the number of ions reaching the mass spectrometer for each detection scan was needed. 42 Analysis of a Mixture of Histones - Experiment 3 Firstly, the column was switched to a pre-made Jupiter C4-packed (5 ȝ m, 300Å, 250x0.3 mm) silica column. Because it is assumed there are more hydrophobic groups on an intact protein than a peptide, the shorter alkyl chains of C4 resin maintain an adequate number of sites for protein interaction while increasing the pore size. Interaction sites within the column provide areas for proteins to differentially partition based on their properties. A larger pore size of 300Å allows easier passage of the intact proteins through the column, which lowers the system pressure and reduces clogging. Using the larger column would fix the pressure issues plaguing the experiment and enable better histone elution. Because of the larger capacity of the column, the flow rate was elevated back to 500 nL/min. Also, the tip was switched back to an unpacked tip until the new column elution could be evaluated. Two other modifications to the system parameters were made to filter out contaminants in the sample: the isolation window value was set to 3.00 and the threshold signal required for the MS 2 scan was set to 50,000. Both of these settings would ensure only the most intense ions of interest were analyzed, thereby focusing and maximizing the detection sensitivity of the mass spectrometer. Lastly, the tube lens voltage was increased to 110 V to guide the highly-charged histone ions into the MS trap. After completion of the mass spectrometer run, histone-like elution envelopes were observable, so MS and MS2 scans were chosen for further analysis. Precursor ion scans at RT 84.44, 85.76, 88.56, and 88.57 were deconvoluted and inputted into ProSight PTM for identification. Similarly, spectra from the product ions produced by the ETD 43 fragmentation of the histones were entered in the ProSight PTM search. The results positively identified all the major histone types and multiple modifications were detected along their sequences (Table 3). Table 3: Histone Matches and Modifications for Experiment 3 Mass (Da) Histone Match Length (a.a.) Mass Difference (Da) Matched Ions Probability Score Modifications 11473.8 Human H4 P62805, P02304, P02305, Histone H4 102 -174.455 16 1.13677e -25 N-acetyl-L- serine N6,N6- dimethyl-L- lysine 14174.8 Human H2A Q96QV6, Q96QV6, BA3 17E16.2 Novel H2A histone family member Histone H2A 131 +38.9205 6 5.52845e -08 N-acetyl-L- serine 14175.7 Human H2B Q8N257, Histone H2B type 12 126 +234.167 3 3.24957e -04 N-acetyl-L- methionine 14293.8 Human H3.4 Q16695, Histone H3.4 H3t H3/t H3/g 135 -1129.69 3 3.24957e -04 N6-methyl-L- lysine N2-acetyl-L- lysine Results from this experiment were promising. Changing the resin of the column to C4 certainly improved the elution of the histones as the back pressure of the system remained at a consistent normal range during the entire run. Even though the certainty was low for a couple of the histone types, there was something to be said about the potential of the workflow that they were identified at all out of the entire known proteome. The low certainty may have been because so few ions were matched to the 44 histone sequences. This would indicate the amount of ETD used for fragmentation is not enough or the amount of each histone precursor being selected for fragmentation is not enough for some of the histone types. Either way, there would not be enough product ions available to match to the histone sequences. Additional improvements to the system parameters would help increase the ETD selection and sensitivity of the histone peaks aiding in the certainty of identification. 45 Analysis of a Mixture of Histones - Experiment 4 Now that the major histone types had been identified by the workflow, the next step was to try and acquire even better certainty of the histone types and confidence in the modifications determined while maintaining machine sensitivity and resolution. Two changes were made to this run of the histone mixture. One change was to add a C18- packed tip (2 cm) to the system to concentrate each histone type before it entered the mass spectrometer. This would help to increase the intensity of the histone elution envelopes thereby triggering more ETD fractionation of the histones and providing more product ions to match to histone sequences and analyze for modifications. Furthermore, the second column-like separation caused by the filled tip helps to increase the distinction between the histone types as they separate. The other change in the workflow was to reduce the amount of histone mixture injected into the machine from 10 ȝ L down to 5 ȝ L. Using a lesser amount of sample would reduce the chance of a clog at the packed tip. Injecting less starting material could potentially drop the intensity of the histone signal, but the small pore size of the resin in the packed tip would compact the more diffusely separated elution fractions from the column before their entrance into the mass spectrometer. Adding the packed tip to concentrate the sample before detection and reducing the sample volume to prevent blockage of the smaller resin pores in the tip would help increase the ETD of more fragment ions for comparison. After the sample had run through the system, the spectra appeared to have areas of histone elution. Precursor spectra were chosen at RT 72.55, 75.08, and, 86.05 for 46 comparison with MS2 spectra generated by the ETD fragmentation. Both sets of spectra were deconvoluted and run in a ProSight PTM search for possible identification and modification characterization. Only a single histone, H4, was identified from the run (Table 4). Several modifications were still detected along the H4 sequence. Table 4: Histone Matches and Modifications for Experiment 4 Mass (Da) Histone Match Length (a.a.) Mass Difference (Da) Matched Ions Probability Score Modifications 9828.76 Human H4 P62805, P02304, P02305, Histone H4 102 -1512.63 6 1.6176e -08 N-acetyl-L- serine N2-acetyl-L- lysine N6,N6- dimethyl-L- lysine These results were a bit of a setback. One explanation for the lack of histones identified from the run was that the injected sample volume decrease did not contain enough histones for analysis. Since histones are found abundantly in the cells, the concentration of the sample was predicted to be at excess and was unlikely the cause of the lack of histones. Conversely, another theory was that the histones compacted too much in the small pores of the C18-filled tip, even with the drop in sample volume, and were not reaching the mass spectrometer for detection. Without elevated signal intensity for the histones by one approach or another, the ETD fragmentation would not be triggered. Also, since the workflow only imputed up to 30 of the most intense peaks for each spectrum into the ProSight PTM search, there may not have been enough adequately intense ETD-produced fragments to be selected for use 47 in the database search. The shortage of product ions would make matching to the precursor ions difficult and would result in a lack of histones identified by the software. Improved intensity of the histone signals would increase the number of ETD fragments produced, leading to more histone identifications. 48 Analysis of a Mixture of Histones - Experiment 5 For this experiment, a single change was made to the workflow to increase the detection of the different histones in a mixture. In order to increase the intensity of the histone peaks during detection in the mass spectrometer, the number of microscans was raised to 10 for the Orbitrap scans in the mass spectrometer. By adding more microscans, each ten microscans in the Orbitrap were summed into one scan before Fourier transformation. Post-FT, the spectra reported represent an accumulation of ten scans. In consequence, the signal intensity of the sample would be increased 10x, which in turn would trigger more ETD fragmentation. From the increase in microscans, more product ions would be generated, providing the ProSight PTM software with better ions to compare to histone sequences. In the spectra generated from the MS run, there emerged numerous possible histone elution envelopes. Chosen from the available elution spectra were the scans at RT 85.70, 87.58, and 90.00. Matching corresponding MS 2 data identified three histone types in the mixture. A few modifications were also found on the histones. The results are summarized in Table 5. Table 5: Histone Matches and Modifications for Experiment 5 Mass (Da) Histone Match Length (a.a.) Mass Difference (Da) Matched Ions Probability Score Modifications 14768.3 Human H2A Q93077, O00775, O00776, O00777, O00778, Histone H2A.l H2A/l 130 +587.367 2 4.33236e -05 N-acetyl-L- serine N2-acetyl-L- lysine 49 Table 5: Histone Matches and Modifications for Experiment 5 (Continued) 14633.3 Human H3 P16106, P02295, P02296, Histone H3.1 H3/a H3/c H3/d H3/f H3/h H3/i H3/j H3/k H3/l 134 -474.032 5 8.21321e -08 None 9412.3 Human H4 P62805, P02304, P02305, Histone H4 102 -1831.06 7 2.53448e -11 N6-methyl-L- lysine An increase in the number of microscans yielded a promising improvement in the successful identification of histones within the sample mixture. The confidence of the identification also improved slightly. However, detection of all the major histone types has not yet been duplicated since the third step of experimentation. Reduction in the overlap between the different histone types could potentially improve the detection and specificity of the analysis. The next step was to try and improve the separation between the histone types to reduce the crossover of eluting histones during detection. 50 Analysis of a Mixture of Histones - Experiment 6 Alterations for this experiment were made to improve the separation between the histone retention times in the column so they would be more differentially detected. One step taken was to increase the flow rate up to 750 nL/min. The purpose of this was to push the histones through the column faster to reduce the diffusion time while the histones separated in the column. Because the sample is carried into the MS system by a liquid vehicle, the separated histones have the opportunity to begin remixing by diffusion with nearby neighbors if they remain in the column too long. Also, the flow increase would improve the histone flow through the pores in the tip. Additionally, the gradient shift between phases was lengthened during the time of past histone elution. This was accomplished by shifting the gradient in smaller increments and over larger periods of time to spread out the elution of the histones generating a more distinct separation between the different types. The lack of overlap made identification of the histone elution envelopes much easier as there would not be any overlapping envelopes to complicate the spectra. A gradient shift would also result in more precise ETD activation because there would not be competing precursors vying for fragmentation. Fragmentation of a more diverse set of precursor ions would lead to more unique fragment ions to be used for histone identification. Following the usual protocol with the improved gradient and faster flow rate, the mass spectrometer returned a few histone envelope candidates. Scans at RT 60.80, 63.46, and 65.44 were selected for MS matching to the MS 2 fragments. Investigation by 51 ProSight PTM concluded three histone types were distinguishable within the chosen spectra and four modifications were detected (Table 6). Table 6: Histone Matches and Modifications for Experiment 6 Mass (Da) Histone Match Length (a.a.) Mass Difference (Da) Matched Ions Probability Score Modifications 13128.2 Human H2B Q8N257, Histone H2B type 12 125 -682.33 3 1.50824e -04 N-acetyl-L- proline 16107.9 Human H3 P16106, P02295, P02296, Histone H3.1 H3/a H3/c H3/d H3/f H3/h H3/i H3/j H3/k H3/l 134 +1000.56 5 8.92161e -08 None 12360.0 Human H4 P62805, P02304, P02305, Histone H4 103 +803.582 3 1.1483e -04 N-acetyl-L- methionine N2-acetyl-L-lysine N6,N6-dimethyl-L- lysine These results did not indicate the gradient changes had a positive effect on the experimental results; in fact, there was a slight decrease in the certainty of the results compared to previous runs. Several experiments were run with varying gradient differences, but all yielded similar insufficient results. Because of the lack of a meaningful increase in the number of product ions formed, mild changes to the gradient were reasoned to be ineffective. One potential explanation for the absence of change was the histones had such slight differences in retention times in this set-up that a gradient shift would not have much of an effect on the separation. Another possibility was the shifts in the gradient were not extreme enough to cause a noticeable alteration in the amount of separation between the histone types. 52 Summary and Conclusions Based on the experimental results, the workflow developed for the study of intact histones correctly identified histone types in a sample and characterized the post- translational modifications along the histone sequences. All four histone types composing the nucleosome octamer were identified and six different PTMs to the histones were characterized by the workflow as shown in Tables 7 and 8 below. Table 7: Summary of Histones Identified in Experiments H2A H2B H3 H4 PTMs Experiment 1 — — YES YES YES Experiment 2 — — — — — Experiment 3 YES YES YES YES YES Experiment 4 — — — YES YES Experiment 5 YES — YES YES YES Experiment 6 — YES YES YES YES Table 8: Summary of Post-Translational Modifications Characterized on the Histone Matches N-acetyl-L- serine N2-acetyl-L- lysine N6-methyl- L-lysine N6,N6- dimethyl-L- lysine Others Experiment 1 YES YES YES YES — Experiment 2 — — — — — Experiment 3 YES YES YES YES N-acetyl-L- methionine Experiment 4 YES YES — YES — Experiment 5 YES YES YES — — Experiment 6 — YES — YES N-acetyl-L- methionine N-acetyl-L- proline 53 While these results are promising and indicate the success of the developed workflow, optimization of additional parameters in the mass spectrometry system is necessary to further increase the sensitivity and accuracy of the results. For example, adjustment of the CI gas pressure would help increase the ETD efficiency to generate additional fragment ions. Fine tuning of the ETD reaction time, automatic gain control, ion transfer capillary temperature, and dynamic exclusion rate would also improve the sensitivity and selectivity of the experiments. The scope of the results obtained from the workflow would expand from optimization of these features, including an increase in identified PTMs. The experiments run provide a framework for future experimentation to be carried out using this workflow. Next steps utilizing this workflow are to further optimize relevant parameters, and then to test different conditional states between sets of histones. Although additional experimentation is needed, the developed protocol for the study of intact histones has already begun yielding information and can be applied as a real-world solution for studying histones. The strength of this workflow lies in its ability to use intact histones while maintaining the integrity of the PTMs for analysis. Utilization of this workflow contributes to the field of epigenetics by providing the means for novel modification mapping, development of epigenetic based therapies for various cancer types, and elucidation of histone modification mechanisms. There is abundant information to be gained from the modification of histones, and the developed workflow adds another method for discovering these changes to increase understanding of the role modifications perform as regulators of cellular events. 54 References Addona T, Clauser K. Current Protocols in Protein Science. 2002; 27:16.11.1-16.11.19. Chi A, Bai DL, Geer LY, Shabanowitz J, Hunt DF. Int. J. Mass Spectrom .2007; 259(1- 3):197–203. Gardner KE, Zhou L, Parra MA, Chen X, Strahl BD. PLoS One. 2011, 6(1, e16244):1-14. Gupta N, Benhamida J, Bhargava V, Goodman D, Kain E, et al. Genome Res. 2008; 18:1133-1142. Hoffert JD, Knepper MA. Anal. Biochem. 2008; 375:1-10. Hu Q, Noll RJ, Li H, Makarov A, Hardman M, Cooks RG. J Mass Spectrom. 2005, 40(4):430-443. Kalli, A, Hess S. Proteomics. 2012, 12:21-31. Li M, Jiang L, Kelleher NL. J Chromatogr B. 2009, 877:3885-3892. ProSight PTM. Kelleher Research Group. Northwestern University. 13 Dec. 2011 <https://prosightptm.northwestern.edu/>. Rabilloud T. Method Mol Bio. 2009, 519:19-30. Roepstorff P, Fohlman J. Biomed. Mass Spectrom. 1984; 11(11):601. Siuti N, Kelleher NL. Nature Methods. 2007, 4(10):817-821. Su X, Lucas DM, Zhang L, Xu H, Zabrouskov V, et al. Proteomics. 2009, 9:1197-1206. Syka JEP, Coon JJ, Schroeder MJ, Shabanowitz J, Hunt DF. P. Natl. Acad. Sci. USA. 2004; 101(26):9528-9533. Yates JR, Ruse CI, Nakorchevsky A. Annu. Rev. Biomed. Eng. 2009; 11:49-79.

Abstract (if available)

Abstract The importance of histones for DNA management cannot be understated. From rearrangement to regulation of expression, histones act as the principle effectors of DNA availability and storage. A majority of diseases involve errors in gene expression

Conceptually similar

PDF

Snake venom proteins identification by nano high performance liquid chromatography tandem mass spectrometry

PDF

Mixed lineage leukemia proteins (MLLs), their effect as coregulators on target gene expression and global histone methylation

PDF

Determinination of the causal potential of histone modifications on transcription and chromatin structure

PDF

Impacts of post-translational modifications on interactions between G9a and its N-terminus binding partners

PDF

Biochemistry and reconstitution of V(D)J recombination in a purified system

PDF

Studies on the role of a novel protein, TMEM 56 in tumorigenic growth for MCF-7 cells

PDF

Identification and characterization of PR-Set7 and histone H4 lysine 20 methylation-associated proteins

PDF

Multiple functions of the PR-Set7 histone methyltransferase: from transcription to the cell cycle

PDF

Studies on the expression and function of the human TMEM56 protein

PDF

The relationship between DNA methylation and transcription factor binding in colon cancer cells

PDF

Characterization of transgenic zebrafish lines for studying role of platelet derived growth factor (PDGF) signaling on coronary vessel formation in regenerating zebrafish heart after cryoinjury

PDF

Characterization of the STARS (STochastic gene Activation with Regulated Sparseness) mouse line

PDF

Functional characterization of colon cancer risk enhancers

PDF

Using epigenetic toggle switches to repress tumor-promoting gene expression

PDF

Characterization of cytochrome c peroxidase of Marinobacter aquaeolei

PDF

Determining the mechanism and necessity of matrix metalloproteinase 2 (MMP-2) nuclear localization for proficient skeletal muscle differentiation

PDF

Alcohol mediated expression of cyto-protective enzyme - NQO-1 and its post translational regulation

PDF

Preosteoblast-specific RUNX2 promotes RANKL membrane association: antagonism by sex steroids through a non-genomic mechanism

PDF

Decreased levels of expression of transmembrane protein 56 (TMEM56) in breast cancer tissues

PDF

The mechanism of R-loop formation in mammalian immunoglobulin class switch recombination

Asset Metadata

Creator Reilly-Rhoten, Heather Noelle (author)

Core Title Identification and characterization of post-translational modifications on histones: an on-line top-down mass spectrometry workflow for analysis using ProSight PTM

School Keck School of Medicine

Degree Master of Science

Degree Program Biochemistry and Molecular Biology

Publication Date 04/19/2012

Defense Date 03/15/2012

Publisher Los Angeles, California (original), University of Southern California (original), University of Southern California. Libraries (digital)

Tag histones,mass spectrometry,OAI-PMH Harvest,post-translational modifications

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Tokes, Zoltan A. (committee chair), Hsieh, Chih-Lin (committee member), Zandi, Ebrahim (committee member)

Creator Email reillyrh@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-8330

Unique identifier UC1111625

Identifier usctheses-c3-8330 (legacy record id)

Legacy Identifier etd-ReillyRhot-620.pdf

Dmrecord 8330

Document Type Thesis

Rights Reilly-Rhoten, Heather Noelle

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA