Virus-producing cells determine the host protein profiles of HIV-1 virion cores

Background Upon HIV entry into target cells, viral cores are released and rearranged into reverse transcription complexes (RTCs), which support reverse transcription and also protect and transport viral cDNA to the site of integration. RTCs are composed of viral and cellular proteins that originate from both target and producer cells, the latter entering the target cell within the viral core. However, the proteome of HIV-1 viral cores in the context of the type of producer cells has not yet been characterized. Results We examined the proteomic profiles of the cores purified from HIV-1 NL4-3 virions assembled in Sup-T1 cells (T lymphocytes), PMA and vitamin D3 activated THP1 (model of macrophages, mMΦ), and non-activated THP1 cells (model of monocytes, mMN) and assessed potential involvement of identified proteins in the early stages of infection using gene ontology information and data from genome-wide screens on proteins important for HIV-1 replication. We identified 202 cellular proteins incorporated in the viral cores (T cells: 125, mMΦ: 110, mMN: 90) with the overlap between these sets limited to 42 proteins. The groups of RNA binding (29), DNA binding (17), cytoskeleton (15), cytoskeleton regulation (21), chaperone (18), vesicular trafficking-associated (12) and ubiquitin-proteasome pathway-associated proteins (9) were most numerous. Cores of the virions from SupT1 cells contained twice as many RNA binding proteins as cores of THP1-derived virus, whereas cores of virions from mMΦ and mMN were enriched in components of cytoskeleton and vesicular transport machinery, most probably due to differences in virion assembly pathways between these cells. Spectra of chaperones, cytoskeletal proteins and ubiquitin-proteasome pathway components were similar between viral cores from different cell types, whereas DNA-binding and especially RNA-binding proteins were highly diverse. Western blot analysis showed that within the group of overlapping proteins, the level of incorporation of some RNA binding (RHA and HELIC2) and DNA binding proteins (MCM5 and Ku80) in the viral cores from T cells was higher than in the cores from both mMΦ and mMN and did not correlate with the abundance of these proteins in virus producing cells. Conclusions Profiles of host proteins packaged in the cores of HIV-1 virions depend on the type of virus producing cell. The pool of proteins present in the cores of all virions is likely to contain factors important for viral functions. Incorporation ratio of certain RNA- and DNA-binding proteins suggests their more efficient, non-random packaging into virions in T cells than in mMΦ and mMN.


Results:
We examined the proteomic profiles of the cores purified from HIV-1 NL4-3 virions assembled in Sup-T1 cells (T lymphocytes), PMA and vitamin D 3 activated THP1 (model of macrophages, mMΦ), and non-activated THP1 cells (model of monocytes, mMN) and assessed potential involvement of identified proteins in the early stages of infection using gene ontology information and data from genome-wide screens on proteins important for HIV-1 replication. We identified 202 cellular proteins incorporated in the viral cores (T cells: 125,mMΦ: 110,mMN: 90) with the overlap between these sets limited to 42 proteins. The groups of RNA binding (29), DNA binding (17), cytoskeleton (15), cytoskeleton regulation (21), chaperone (18), vesicular trafficking-associated (12) and ubiquitinproteasome pathway-associated proteins (9) were most numerous. Cores of the virions from SupT1 cells contained twice as many RNA binding proteins as cores of THP1-derived virus, whereas cores of virions from mMΦ and mMN were enriched in components of cytoskeleton and vesicular transport machinery, most probably due to differences in virion assembly pathways between these cells. Spectra of chaperones, cytoskeletal proteins and ubiquitinproteasome pathway components were similar between viral cores from different cell types, whereas DNA-binding and especially RNA-binding proteins were highly diverse. Western blot analysis showed that within the group of overlapping proteins, the level of incorporation of some RNA binding (RHA and HELIC2) and DNA binding proteins (MCM5 and Ku80) in the viral cores from T cells was higher than in the cores from both mMΦ and mMN and did not correlate with the abundance of these proteins in virus producing cells. Conclusions: Profiles of host proteins packaged in the cores of HIV-1 virions depend on the type of virus producing cell. The pool of proteins present in the cores of all virions is likely to contain factors important for viral functions. Incorporation ratio of certain RNA-and DNA-binding proteins suggests their more efficient, non-random packaging into virions in T cells than in mMΦ and mMN.

Background
HIV-1 viral particles released from infected cells have been shown to incorporate many cellular proteins during the assembly and budding steps of morphogenesis. Findings from earlier studies, summarized in a web-based database (http://web.ncifcrf.gov/research/avp/protein_db. asp), identified more than three hundred cellular proteins in HIV-1 particles. HIV-1, as well as other lentiviruses, incorporates components of the cellular endosomal sorting machinery and cytoskeleton proteins involved in the process of particle assembly [1][2][3][4][5], surface proteins captured with the plasma membrane during budding [6,7], RNA-binding proteins associated with incorporated viral RNA and RNA-Gag complexes [4,[8][9][10], chaperones [11], and multiple concomitant proteins (reviewed in [12]) whose functions in viral morphogenesis and infectivity are still unknown.
Upon fusion of an HIV-1 particle with a target cell, viral cores are released into the cytoplasm and rearranged into sub-viral particles called reverse transcription complexes (RTCs), which subsequently mature into pre-integration complexes (PICs). These nucleoprotein structures support reverse transcription and also protect and transport viral cDNA to the site of integration. RTCs are composed of both viral and cellular proteins. Since RTCs are formed from the viral cores, their initial composition is identical to that of viral cores. Other than the key enzymatic components, reverse transcriptase (RT) and integrase (IN), at least five other viral proteins involved in structural organization, cytoplasmic trafficking and nuclear import (matrix [MA], nucleocapsid [NC], capsid [CA], Nef and viral protein R [Vpr]), have been identified as components of HIV-1 RTCs [13][14][15][16][17][18][19], reviewed in [20][21][22].
Although the key early steps of HIV replication, reverse transcription and integration, are relatively autonomic, the participation of cellular proteins in early infection events has been demonstrated in previous studies [23][24][25]. After release from the viral particles, RTCs are still encapsulated in the shells formed by p24 CA molecules which are stable in the cytoplasm for at least several hours [26,27]. The shell is believed to protect the reverse transcription machinery and all encapsulated proteins from the cytoplasmic environment to provide optimal conditions for their functional activity [28] and may contribute to the nuclear import of PICs [19,[29][30][31]. The shell likely limits the access of host cell proteins to the RTC interior. Thus, most cellular factors which may contribute to the functional competence of early RTCs should be expected to get into the complexes from the cores of infecting virions.
The cellular proteins, which are known to be hijacked by assembling virus particles from virus-producing cells and are involved in the early post-entry stages of HIV-1 infection, can be grouped into the following categories.
(1) Factors involved in the spatial organization and correct folding of viral proteins in the virion and probably RTC: clathrin [5,32] and heat shock proteins (Hsp70, Hsc70, Hsp60) [11,33] are probably critical for spatial organization of Gag and Pol proteins and regulation of proteolytic processing and folding of the Pol products -RT and integrase; thioltransferase is found in HIV-1 virions and may be important for dimerization and activation of the viral protease [34], (reviewed in [12]); staufen1, an RNA-binding protein is packaged into virions and is involved in incorporation of HIV-1 RNA [8]. Interaction of staufen1 with Pr55 Gag zing finger motifs may also be important for Gag multimerization and formation of the viral capsids [35]. (2) Proteins which have an effect on cDNA synthesis/accumulation: lysyl-tRNA synthetase is incorporated through the interaction with Gag and is critical for the priming of reverse transcription [36]; uracil DNA glycosylase 2 (UNG2), a cellular DNA repair enzyme that binds HIV-1 integrase and Vpr [37,38] . The role of this enzyme in early post-entry steps of infection remains controversial. The hypothesis that the catalytic activity of Vpr-associated UNG could modulate virus mutation rate and APOBEC3G-mediated G-to-A hypermutations [39][40][41] was not supported by subsequent studies [42]. However, recently published work of Guenzel and co-authors showed that the virionincorporated nuclear form of UNG2 facilitated reverse transcription through a non-enzymatic mechanism involving direct interaction with the p32 subunit of the replication protein A (RPA) complex [43]. RNA helicase A (RHA or DHX9) is packaged into HIV-1 virions probably through the interaction with an RNA or Gag polyprotein and facilitates reverse transcription [10]. The protein INI1/hSNF5, a member of SWI/SNF chromatin remodeling complex, has been shown to be packaged into virions through the direct binding to integrase [44], and is involved in the synthesis of reverse transcription products [45]. Later studies demonstrated that INI1/ hSNF5 selectively recruits into HIV-1 virions the components of Sin3a-HDAC1 cellular complex, whose presence is critical for the early reverse transcription stage [46]. Furthermore, interaction of HIV-1 integrase with INI1 has been shown to be essential for the nucleosome remodeling of host chromatin and hence overcoming the structural nucleosome barrier for viral integration [47]. (3) Proteins involved in RTC formation, protection and transport: cyclophilin A, which is incorporated into virions via binding to the CA domain of Pr55Gag [23]. The role of CA-bound cyclophilin A in the viral life cycle is still unclear [48] Recently published data showed this protein to be critical for protection and stabilization of HIV-1 cores [49]. It may also be involved in PIC nuclear transport [31]. (4) Restriction factors of the early stages of HIV-1 infection: members of the APOBEC3 family of DNA/RNA editing cytidine deaminases, APO-BEC3G (A3G) and APOBEC3F (A3F), are incorporated in Vif-negative virus particles (a small amount of these factors may be present also in Vif-positive virions) and then restrict reverse transcription by carrying out hypermutation of newly synthesized HIV-1 DNA [41,[50][51][52]. Numerous studies have shown that A3G molecules from the target cell have no effect on cDNA deamination, and only virion-incorporated A3G affects viral DNA (reviewed in [53]), suggesting that the cDNA synthesis and accumulation machineries are effectively isolated from the environment of the target cell cytoplasm, but can be affected by factors which are encapsulated in cores and found within RTCs. Initiation of uncoating or disintegration of the capsid shell is believed to be dependent on the completion of reverse transcription [27,28]. Uncoated complexes containing viral cDNA are capable of interacting with numerous factors of the target cell, which are necessary for nuclear import and probably facilitation of viral genome integration (reviewed [21,54]).
Although numerous studies, including recently published proteomic analyses of whole highly-purified retroviral virions or virus-like particles, contain broad information about profiles of cellular proteins in viral particles [4,10,55,56], the host proteins associated specifically with the cores of mature HIV-1 virions have not been characterized. To identify the proteomic profiles of HIV-1 cores , we performed LC-MS/MS analysis of the cores isolated from virus particles of HIV-1 NL4-3 produced by Sup-T1 cells (T lymphocytes), PMA and vitamin D 3 activated THP1 (model of monocyte-derived macrophages), and non-activated THP1 cells (model of monocytes), and Western blot analysis of selected proteins in the cores and producer cells. Potential involvement of identified proteins in the early stages of HIV-1 infection was assessed using gene ontology information and data from published genome-wide screens on proteins important for HIV-1 replication [57][58][59]. Our study revealed 202 proteins associated with HIV-1 cores. More than 20% of these proteins were detected in the cores of virions from all cell types, suggesting that this group contains cellular proteins potentially involved in viral replication. We found that some members of this group, which belong to subfamilies of RNA and DNA helicases, are packaged into the virions from producing T cells more efficiently than into virions from monocyte or MDM model cells, indicating that the mechanism of their incorporation is nonrandom.

Preparation of HIV-1 core structures for proteomic analysis
Since HIV-1 is a highly variable virus, its different subtypes and individual variations may interact differently with cellular proteins [60]. Therefore, we selected the NL4-3 isolate of HIV-1 subtype B as a model virus for infection of different cell types: T lymphocytes and the model of monocytes and MDM. NL4-3 is a CXCR4tropic isolate of HIV-1 that normally does not infect monocytes and macrophages, so to infect THP1 cells we used virus pseudotyped with amphotropic murine leukemia virus (MLV) Env glycoprotein. For isolation of viral cores and subsequent LC-MS/MS analysis, we used the virus particles produced by infected Sup-T1 cells (T lymphocyte cell line), THP1 (monocytic leukemia cells) and THP1 cells treated with phorbol 12-myristate 13acetate (PMA) and vitamin D 3 . This treatment activates cell differentiation that results in acquisition of the biochemical and morphological characteristics of MDM.
Thus, the activated THP1 cells may be considered as a model of macrophages [61,62].
To isolate core structures from HIV-1 virions produced by infected cells, we engaged a technique of "spin-thru" equilibrium density gradient sedimentation described earlier [63][64][65][66]. This method of ultra-speed centrifugation of previously concentrated HIV-1 virions through a sucrose density gradient overlaid with a detergent layer (1% Triton X-100) allows for the purification of mature lentiviral cores whose density varies from 1.23 to 1.27 g/ml [67,68] ( Figure 1A, lower panel), whereas intact viral particles display buoyant density 1.18-1.20 mg/ml ( Figure 1A, upper panel). To establish the purity of our viral core preparations from cellular vesicles, which have density similar to that of virions (1.14-1.20 g/ml) and may contaminate virus preparations [69], and to compare maturation of capsid cores in the viruses produced by T lymphocytes and MDM model cells, we engaged electron microscopy and Western blot analysis. Examination of the negatively stained concentrated virion samples in a transmission electron microscope revealed the presence of both extracellular vesicles and viral particles with diameter from 120 to 130 nm ( Figure 1B1-B3). Analysis of ultrathin sections of the viral particles used for core isolation showed that the population represented a mix of immature ( Figure 1B4, black double arrows) and mature virions ( Figure 1B4, single arrow). The lipid membrane-covered structures were not found in purified viral core preparations. Only the conical-shape and oval structures with length about 80-110 nm were found in the samples of purified viral cores, indicating that the preparations after "spin-thru" purification contained only mature capsid cores ( Figure 1B5-B8). Previous studies of cores from HIV-1 [64] and HIV-2 virions [63] isolated by "spin-thru" centrifugation method showed the presence of mature products of Gag and GagPol proteolytic processing (CA, MA, Vpx [for HIV-2] and RT with high enzymatic activity). Immunoblotting of our core preparations obtained after "spin-thru" purification of the same amounts of Sup-T1-and THP1-derived viral particles, carried out with human IgG prepared from pooled plasma of HIV antibody positive donors, revealed similar amounts of major products of Gag and GagPol processing, such as MA, CA, IN and RT proteins ( Figure 1C). The Pr55 Gag and Pr41 Gag (MA + CA) have also been identified in the core samples from both viruses, suggesting that mature cores may contain some amounts of unprocessed Gag polyprotein. Taken together, our data indicate that the selected method of purification allows for the isolation of mostly mature cores from the pools of virions produced by both T lymphocytes and MDM model cells.
We also performed Western blotting of our samples using the anti-CD45 antibody (Figure 2A). CD45 is known to be abundant in microvesicles, but is apparently excluded from HIV-1 virions [70]; the lack of this protein in our core preparations would confirm their purity from the vesicular fraction. Indeed, we did not detect CD45 in the samples of cores from HIV-1 virions produced by both Sup-T1 and THP1 cells (Figure 2A, two right lanes), whereas the specimens of culture media from these cells concentrated only through a 30% sucrose cushion contained detectable amounts of CD45. Western blotting of the samples of culture media from untransfected and NL4-3 proviral clone-transfected 293 T/17 cells using anti-RNA helicase A (RHA or DHX9) antibody showed presence of this DEAD box RNA helicase in the preparations of media from both transfected and untransfected cells after purification through 30% sucrose. However, in the "spin-thru"    Figure 2B). Since RHA is known to be present in both vesicles [71] and HIV-1 virions [10,72,73], our analysis confirmed that the method of "spin-thru" centrifugation removed extracellular membranous structures from the 30% sucroseconcentrated cell culture supernatants, but retained intravirion core structures.
To further prove that "spin-thru" centrifugation purifies cores from intact virions, we tested the presence of VSV-G envelope protein in the samples of VSV-Gpseudotyped HIV-1 produced by co-transfected 293 T/ 17 cells after concentration through 30% sucrose cushion and "spin-thru" centrifugation ( Figure 2C). The VSV-G was clearly detected in the samples of concentrated pseudotyped virus, but was not found in the core samples after "spin-thru" purification, confirming purity of the core preparations from the envelope glycoproteins.
The SDS-PAGE separation of our core preparations ( Figure 2D) revealed major bands corresponding to proteins with molecular weights of 24 and 31 kDa (corresponding to HIV-1 CA and IN, respectively), indicating the presence of mature viral Gag and GagPol products in the analyzed core structures. On the other hand, multiple bands corresponding to the polypeptides of different molecular weights, which do not represent known HIV-1 proteins, suggest incorporation of many cellular proteins in the core structures of the viral particles produced by different cell types. The data of proteomic analysis (shown below) confirmed this suggestion. The staining of SDS-PAGE with Coomassie also revealed  multiple protein bands in the control preparations, suggesting that the culture media from uninfected THP1 and especially Sup-T1 cells contained protein-rich, nonviral, non-membranous particles with buoyant density ≥1.23 mg/ml, probably the products of disintegrated dead cells ( Figure 2D). Thus, to obtain proteomic profiles of the host proteins associated with HIV-1 viral cores, both viral cores and uninfected control preparations from each cell type were subjected to SDS-PAGE protein separation, trypsin digestion and subsequent LC-MS/MS analysis. The protein profile of each viral core sample was then compared with the corresponding control sample. Overlapping proteins were eliminated from the protein spectra of the viral cores, except the proteins whose scores were >5-fold higher in the preparations of viral cores than in control samples (proteins such as chaperones Hsp70 and Hsp90, and cytoskeletal proteins β actin, α and β tubulin, whose presence in core samples was confirmed by Western blot [ Figure 4A]). As a result, a total of 202 cellular proteins were found to be associated with the cores of HIV-1 virions.

Proteomic profiling of HIV-1 viral cores
The proteins obtained from LC-MS/MS analysis of peptide preparations and filtered as described above and in Materials and Methods were categorized according to their functions and subcellular localization using NCBI protein database (http://www.ncbi.nlm.nih.gov/sites/entrez? db=Protein), NCBI RefSeq database (www.ncbi.nlm.nih. gov/RefSeq/) and DAVID Bioinformatics Resources 6.7 (NIAID NIH) (http://david.abcc.ncifcrf.gov). We compared our data with published results of the genome-wide analyses of cellular proteins involved in HIV-1 infection [57][58][59]74] and with global network of HIV-human protein-protein interactions [75] to assess the potential role of identified proteins in HIV-1 infection and putative mechanisms of their incorporation into the virion. The HIV-1 proteins identified in the viral cores are summarized in Table 1, the cellular proteinsin Tables 2 and S1 in Additional file 1.

Viral proteins
The major proteins constituting the HIV-1 nucleocapsid core, CA, NC, IN and RT (both p51 and p66 subunits) were identified in the preparations from virions generated by all cell types (Table 1). All core samples also contained MA protein, which is located in retroviral virions mostly outside the capsid and forms a matrix between the viral capsid and envelope [76,77]. However, presence of MA in the RTCs and PICs [13,14,17,78,79] suggests that this protein is physically associated with the cores of HIV-1 virions. Comparison of our list of identified viral proteins with the MS/MS analysis data of whole MDM-produced HIV-1 particles [4] shows characteristic differences between the protein profiles of viral cores and whole virions. The gp120 Env glycoprotein detected in whole virions is absent in our samples; however, the gp41 transmembrane (TM) Env product was identified in the cores from virions assembled in T lymphocytes and MDM-like activated THP1 cells. Since gp41 has been found to be associated with MA during virion assembly and probably in mature viral particles [80,81] (reviewed in [82]), a low amount of this glycoprotein in detergent-purified core preparations was expected. The fact that we could not identify gp41 in the cores of virions from non-activated THP1 cells by LC-MS/MS confirms our suggestion that the concentration of this glycoprotein in our preparations is negligible. The absence of gp120 Env in our samples confirms purity of isolated core structures from the viral envelope and intact virions, shown also in Figure 2C. The Tat, Nef and Vif proteins were identified earlier in the viral particles produced by infected MDM using proteomic methods [4]. We did not detect Tat in our core preparations, but MS/MS analysis revealed Nef in the viral cores from all analyzed cell types. Earlier studies indicated that Nef is incorporated into viral particles, stably associates with virion cores and facilitates reverse transcription and early steps of replication [16,64,83,84]. The Vif and Rev proteins were identified with low scores only in the cores from T lymphocyte-derived virus, suggesting their very low concentrations in the cores. HIV-1 Vif has been shown to interact with Pr55 Gag and viral protease during the assembly of virus particles [85]. Rev protein can be incorporated into cores in the complex with viral RNA. Surprisingly, we did not identify Vpr in the viral core samples. Numerous studies revealed Vpr in HIV-1 virions [86][87][88] and RTCs [15,[89][90][91]. However, while traditional methods of identification using specific antibodies recognize Vpr in HIV-1 particles, LC-MS/MS analysis performed by Chertova and co-authors [4] also did not reveal this protein in the preparations of whole highly-purified virions. The visibility of protein for LC-MS is defined by a few factors, including affinity to C18 column (hydrophobicity) and ionization efficiency of the peptides. Typically, about 10% of tryptic peptides of a long protein are visible. Thus, some proteins could be invisible for LC-MS/MS in spite of overall high sensitivity of a method. Moreover, although trypsin cleaves Vpr to 14 peptides, most of them are not charged or have a negative total charge in solution with neutral pH (analyzed using Innovagen Peptide property calculator tool [http://www.innovagen.se/]), which may cause additional problem with their detection by mass spectrometry.
HIV-1 protease (PR) was observed in all our core preparations. At the same time, detection of Gag and Gag-Pol polyprotein-precursors in respective gel fractions suggests that not all precursor molecules are subjected

Cores of virions assembled in:
Protein categories (for C1-C4) to the proteolytic cleavage during the virion assembly and a subset of unprocessed polyproteins is present in the mature viral cores.

Cellular proteins
Within 202 unique cellular proteins revealed in our preparations of purified cores from HIV-1 virions, the samples from Sup-T1-derived virus included 125 proteins, while the virion cores from activated and non-activated THP1 cells contained 110 and 90 proteins, respectively ( Table 2). A similar number of common proteins was found between cores from viruses produced by activated and non-activated THP1 cells (51), non-activated THP1 and Sup-T1 cells (63), and activated THP1 and Sup-T1 (52) ( Figure 3A). Forty two proteins were common to all viral cores, which equates to 34%, 38% and 47% of the proteins found in the cores of viruses derived from Sup-T1, activated and nonactivated THP1 cells, respectively.
We identified 125 host cell proteins in the cores of virions assembled in Sup-T1 cells; cores of the viruses from activated THP1 cells contained 12% less host proteins, whereas the number of host proteins in the cores of non-activated THP1-derived virus was 28% less than in the cores of virus from Sup-T1. Interestingly, only 20% of all cellular proteins identified in the cores of virus from non-activated monocytic cells were unique, other proteins were also found in the cores of the virions derived from Sup-T1 or activated THP1 cells. The cores of activated THP1 and Sup-T1-derived viruses, in contrast, contained 45% and 42% of unique host proteins, respectively ( Figure 3A). Since the T lymphocytes and MDM (activated THP1 is a model of MDM) naturally (See figure on previous page.) Figure 3 Profiles of host proteins associated with the cores of HIV-1 virions from different producer cellsdata of LC-MS/MS analysis. A, B -Venn Diagrams depicting the number of overlapping cellular proteins within viral cores among the three producer cell types (A) and within the cores of viruses produced by the same cell type (Sup-T1) infected with HIV-1 NL4-3 strain pseudotyped with MLV Env (yellow) or VSV-G (green) envelope glycoproteins (B). The number of non-overlapping proteins and percent of these proteins within all cellular proteins identified in the core are shown as a numerator and denominator respectively. C -Categories of cellular proteins (by function) depicted as a percentage of the total proteins that were identified within the cores of virions produced by Sup-T1 (C1, n = 125), THP1 cells activated with PMA and vitamin D 3 (C2, n = 110), and non-activated THP1 cells (C3, n = 90). The diagram of overlapping cellular proteins detected in all core preparations is shown on C4 (n = 42).         [92]), the incorporation of a smaller number of cellular proteins by the virions released from non-activated THP1 is probably related to the lower efficiency of virion assembly in this type of cells.
To validate the credibility of differences in protein composition between viral cores from different cell lines, we performed proteomic analysis of two core preparations purified from the same cell line. Cores of virions harvested from Sup-T1 cells infected with MLV Envpseudotyped and VSV-G-pseudotyped NL4-3 viruses were digested by trypsin without gel separation and subjected to LC-MS/MS. Per one hundred high scoring overlapping proteins we found only 18 unique cellular proteins incorporated in the viral cores from the MLV Env-pseudotyped virus-infected Sup-T1 (15% of all incorporated host proteins) and 10 proteins (9%) in cores of the virus from Sup-T1 cells infected with VSV-G-pseudotyped NL4-3 strain ( Figure 3B; Additional file 1: Table S1). Since the minimal difference in the spectra of host proteins between cores of virions harvested from different cell lines was not less than 20%, and was equal to 42% and 45% for the viral cores from Sup-T1 and activated THP1 cells, respectively, relative to cores from non-activated THP1 cells, we believe that the detected differences between the profiles of cellular proteins in the viral cores from various cell lines are significant.
All the cellular proteins in HIV-1 cores were classified into 31 functional categories ( Table 2). The most numerous categories were as follows: RNA-binding proteins (29), components of the cytoskeleton (15) and cytoskeleton regulators (21), chaperones (18), DNA-binding proteins (17), proteins involved in vesicular transport (12) and components of the ubiquitin-proteasome system (9). Although these protein groups were numerous in all preparations, the viruses assembled in different types of producer cells demonstrated diversity in the protein spectra and number of proteins within each group ( Figure 3C).
Within the forty two cellular proteins which were present in all core preparations, the spectra of molecular chaperones (7), cytoskeleton components (7), and vesicular transport-associated proteins (5) were the most numerous ( Figure 3C4). These functional groups of proteins have been shown to be involved in the folding of viral proteins and HIV-1 virion assembly (reviewed in [12,93]). Some of these proteins (clathrin, transferrin receptor 1, RAB7, RAB5C, EHD4, Hsp70, Hsp90, cyclophilin A, β actin, tubulin α1) are very typical for HIV-1 virions and were registered earlier in the samples of purified viral particles (summarized in the database of Host Proteins in HIV-1 [http://web.ncifcrf.gov/research/ avp/protein_db.asp]). Probably some of these proteins, such as β actin, α and β tubulin, moesin and major vault protein 1 are incorporated into HIV-1 virions non-specifically, due to their close proximity to a budding site, as suggested earlier [4,94], whereas other cellular proteins may be incorporated due to specific interactions with viral proteins, such as Gag and GagPol [11,23,33], viral genomic RNA [8], or tRNA Lys3 primer [73]. These host proteins, called "Captives" in a recent review [12], may be involved in the virion assembly and budding process or be important for post-assembly steps of HIV-1 life cycle. Below we attempt to assess the packaging of selected cellular proteins in the cores of HIV-1 virions depending on the type of producer cells.

Semi-quantitative analysis of selected cellular proteins packaging into the cores of HIV-1 virions assembled in different types of cells
Although the spectra of the cellular proteins incorporated in the viral cores depended on the type of virusproducing cells, the quantitative differences within the group of overlapping proteins could also be observed. We compared the abundance of certain cellular proteins, selected from the group of 42 common proteins, in the viral cores and lysates of the producer cells. Since the group of 42 cellular proteins identified in all viral cores included the members which are likely involved in viral replication (Table 3), we selected 7 proteins from  different functional categories for analysis by Western blot. Within the group of RNA-binding proteins, we analyzed incorporation of RNA helicase A, because the function of this protein in HIV-1 replication has been shown earlier [10,95], and small nuclear ribonucleoprotein 200 kDa (U5) (SNRNP200 or HELIC2), which currently has no known role in HIV replication but has been detected with a high score in our MS/MS preparations. Within DNA-binding group of proteins, we analyzed the Minichromosome maintenance complex component 5 (MCM5) involved in initiation of DNA replication, 80-kilodalton subunit of the Ku heterodimer protein or ATP-dependant DNA helicase II (Ku80 or XRCC5), which is involved in the repair of DNA doublestrand breaks and telomerase function [96,97], Pontin52 (RUVBL1) and Reptin52 (RUVBL2) DNA helicases, both being the components of several high molecular weight protein complexes involved in chromatin remodeling, transcription regulation, DNA damage sensing and repair [98]. All these DNA-binding proteins were detected with high scores by LC-MS/MS in the cores of virions from both Sup-T1 and THP1 cells. Beta tubulin was selected as one of the major cytoskeletal proteins found earlier in HIV-1 particles (http://web.ncifcrf.gov/research/avp/protein_db.asp). All preparations were normalized to CA p24 Gag ; the cell lysates were additionally normalized by the cell count and β globin DNA count using quantitative real-time PCR. Cytoskeletal proteins, such as actin or β tubulin, could not be used for normalization of the cell lysates, because they are differently expressed in uninfected Sup-T1 and THP1 cells ( Figure 4A). Analysis of selected proteins using Western blot with subsequent quantification of the band intensity using ImageJ software did not reveal differences in the abundance of β tubulin between the infected cells and the cores of virions produced by these cells (Figure 4B,C). The same ratio of protein concentrations between cells and viral cores was also detected for RUVBL1 and RUVBL2 DNA helicases. However, other analyzed DNAand RNA-binding proteins demonstrated either moderate (XRCC5, SNRNP200) or high prevalence (MCM5, DHX9) in the cores of virions assembled in Sup-T1 cells, whereas in the cores of THP1-derived virions (both activated and non-activated) these proteins were less abundant. Incorporation of these proteins in the viral cores did not correlate with their abundance in the producer cells: XRCC5, MCM5 and SNRNP200 were presented in all types of cells at a similar level, whereas the RNA helicase A (DHX9) was abundant in dividing Sup-T1 and non-activated THP1 cells and decreased in non-dividing activated THP1 ( Figure 4B,C). Interestingly, the incorporation of RNA helicases DHX9 and SNRNP200 into the cores did not correlate with the packaging of viral genomic RNA, as quantitative RT-PCR of the RNA isolated from CA p24 Gag -normalized purified viral cores revealed no significant differences in the count of HIV-1 RNA in the cores from Sup-T1 and both THP1-derived viruses ( Figure 4D).

Discussion
The proteomic analysis of the cores isolated from HIV-1 virions assembled in T lymphocytes and activated (model of MDM) and non-activated monocytic cells revealed more than thirty cellular proteins which have been previously shown to be involved in different steps of HIV-1 replication and/or incorporated in HIV-1 virions (Table 3). Since HIV-1 is a retrovirus with a RNA genome of less than 10 kB and encodes only nine polypeptides, it engages numerous cellular factors and pathways at all stages of its life cycle. Some of the factors, especially proteins involved in RNA splicing and nuclear export, multi-vesicular bodies (MVB) and late endosomal pathway, as well as the proteins directly involved in the HIV-1 budding process, such as AIP1/Alix, may be incorporated into virions by being associated with the viral RNA and proteins, but do not play a visible role in subsequent stages of viral replication [4,125,130]. For (See figure on previous page.) Figure 4 Incorporation of certain RNA-and DNA-binding cellular proteins into HIV-1 viral cores does not correlate with abundance of these proteins in infected cells. A -Western blot detection of the cytoskeleton proteins actin and β tubulin in uninfected Sup-T1, activated and non-activated THP1 cells. Lysates were normalized according to cell counts and then according to the count of β globin DNA using quantitative real-time PCR, and subjected to SDS-PAGE and Western blot analysis. B -Western blot detection of cellular RNA-binding proteins (DHX9, SNRNP200), DNA-binding proteins (MCM5, XRCC5, RUVBL1, RUVBL2), cytoskeleton protein β tubulin, and viral protein CA p24 Gag in the lysates of virus-producing cells (left bands) and in "spin-thru" purified viral cores (right bands). Virus was harvested at 72 h p.i. from Sup-T1, activated and non-activated THP1 cells infected with MLV Env-pseudotyped HIV-1 NL4-3, normalized to CA p24 Gag and subjected to the "spinthru" core isolation. Lysates of infected cells were normalized according to total protein count and β globin DNA count as described in A. Cellular and viral core preparations were analyzed by Western blotting. C -Quantification of Western blotting results. Western blotting data were quantified using ImageJ software. Results are presented as percentage of the peak value for each protein in the cellular and viral core preparations. D -Quantification of viral genomic RNA in the cores of virions. Viral cores were prepared as described in Figure 1. RNA was isolated from CA p24 Gag -normalized core samples, subjected to reverse transcription with oligo-dT primer and then to quantitative real-time PCR with the primer set specific for positive-strand HIV-1 DNA. The data represents analysis of three independent preparations. Each point shows mean RNA copy number ± SD per 1 ng of p24 CA in the viral core sample. 3.32 1.78 -RCC proteins interact with Rac1 and Arf6 subnetworks and limit signaling required for membrane protrusion and delivery [99]; RCC2 acts as a Rac1 guanine nucleotide exchange factor (GEF); RCC1 is involved in HIV-1 RNA nuclear export through activation of RanGAP [100] and hence facilitates dissociation of RNA nuclear export complex [101]. ; may also be involved in expression and nuclear export of retroviral RNAs, particularly in post-transcriptional regulation of HIV-1 [104]. DHX9 is packaged in HIV-1 particles and contributes to particle assembly and reverse transcription [10]; it also facilitates tRNA Lys3 binding and initiation of reverse transcription [73] Heterogeneous nuclear ribonucleoprotein A1 HNRNPA1-1.92 -RNA binding proteins that complex with heterogeneous nuclear RNA (hnRNA) and are involved in pre-mRNA processing in the nucleus: alternative splicing regulation, polyadenylation, nucleo-cytoplasmic transport and other aspects of mRNA metabolism and transport; hnRNP-A1 is involved in HIV-1 mRNA splicing [105,106]; hnRNP-A2 is found to be important for trafficking of HIV-1 mRNA out of the nucleus and through the cytoplasm [105]; hnRNP H and hnRNP K interact directly with HIV-1 RNA and are involved in alternative splicing [107].  [112] and transport of RTC toward the nucleus [18].
the other factors, such as Hsp70, CLTC protein/clathrin and Rab GTP proteases, their important role in the molecular organization of mature virions and probably viral entry into the target cells has been proposed earlier [5,11,32,33,58]. Here, we focus on the proteins identified in the core structures of HIV-1 virions assembled in different cell types. These proteins can be potentially involved in post-entry stages of the viral replication. The HIV-1 morphogenesis is known to be different in T lymphocytes and myeloid cells. In T cells, the viral particle budding and assembly have been shown to take place directly at the plasma membrane (reviewed in [131]), whereas in macrophages, earlier studies detected assembling HIV-1 particles in the late endosomes [132] (reviewed in [131]) or in internally sequestered plasma membrane domains that contain late endosomal markers but are connected to the cell surface [133]. Recent reports revealed an extensive tubular network and large sheet-like structures which extended to the cell surface from vesicular compartments and contained HIV-1 Table 3 Previously discovered viral proteins with a known role in HIV-1 Replication a (Continued) Nuclear Import Transportin 3; importin-SR; importin 12; transportin SR2 TNPO3 1.0 --Impotin-β family member, binds catalytic core domain close to the N terminus of IN and promotes nuclear entry of PICs [115,116]; might serve as a chaperone that associates with PIC post-entry to guide it through nuclear pore [117].

Vesicular Transport
CLTC protein -clathrin CLTC 37.6 7.7 7 Clathrin is incorporated in HIV-1 particles probably through interaction with Pol, especially IN domain [32]; it facilitates the accurate morphogenesis of infectious particles probably by contribution to spatial organization of Gag and Pol proteins and proteolytic processing of virion components during particle assembly [5].
Rab5C GTP protease RAB5C -6.78 6.58 Rab GTP proteases are important for vesicular trafficking. They are activated by guanine nucleotide exchange factor (GEF), RCC2 protein revealed in HIV-1 cores can act as a GEF. Rab11 is important for HIV-1 production [123]; Rab1 potentially associates with HIV-1 Rev and is involved in nuclear export of viral RNA [124]; Rab9 is required for Gag trafficking to the site of assembly [123]; Rab7-interacting lysosomal protein promotes vRNA clustering at the MTOC [112]; Rab6 is probably involved in viral entry [58].

Chaperones/Molecular Folding
Hsp70 protein 8 HSPA8 7. 85 26.4 31.35 Heat shock protein 70 family members are shown to be incorporated in HIV-1 particles. This is important for subsequent viral cDNA synthesis [11,127]; they can also interfere with Vpr in HIV-1 nuclear import in macrophages [128,129]. Hsp70 protein 9 (mortalin) HSPA9 -5.75 -Cyclophilin A, CyPA PPIA 22.3520.00 7.68 incorporates into virions via binding to the CA domain of Pr55Gag [23]. The role of CA-bound CyPA is still unclear [48] It is critical for protection and stabilization of HIV-1 cores as a chaperone [49] and is probably involved in PIC nuclear transport [31] a Proteins are listed by function category. b Full Protein name as listed in NCBI Protein database (http://www.ncbi.nlm.nih.gov/protein). c Official Gene Symbol as listed by HGNC. d The Xcorr values of each protein in different viral core samples are shown; if protein was not present in the virion core from a specific producer cell type, the Xcorr is not shown. e Protein function within the cell is listed, along with known implicated function(s) in HIV-1 replication.
virions, released into the extracellular media [134,135]. Indeed, the cores of model MDM-derived virions contained twelve vesicular trafficking-associated proteins, whereas only six were detected in the T cell-derived viral cores. The larger proportion of cytoskeleton and cytoskeleton regulatory proteins in the cores of virions assembled in THP1 cells than in Sup-T1-derived cores may be dependent on the abundance of these proteins in producer cells. Uninfected THP1 cells contain larger amounts of actin and β tubulin than Sup-T1 ( Figure 3A), suggesting that other cytoskeleton and associated cytoskeleton regulatory proteins may also be more abundant in these myeloid cells; hence, increasing the probability that the virus will hijack this subset of proteins. Our data suggest that the unique core-incorporated proteins, which are different in the viruses assembled in different cell types, are mostly indiscriminately hijacked during virion assembly and likely not important for subsequent infection. The group of forty two cellular proteins identified in the cores of virions produced by all types of cells contains at least thirteen proteins whose involvement in different stages of HIV-1 infection has been shown previously (Table 2 and 3). On the other hand, many proteins within this group have never been found to be implicated in any infection event. Meanwhile, incorporation of these proteins into the viral cores from different types of producer cells suggests that at least some of them may be important for successful infection. For instance, within the functional category of vesicular trafficking-associated proteins, the member of ESCRT pathway AIP1/ALIX detected in all our core preparations has been shown earlier to be interacting with HIV-1 Gag late domain and to be important for the release of viral particles [125,136]. Clathrin has also been found to be abundant in HIV-1 viral particles and important for the correct assembly and maturation of viral particles through the regulation of proteolytic processing of virion components [5,32]. Members of the RAB family of proteins were found to be important for different steps of HIV-1 particle assembly and probably RNA incorporation [112,123,124] (Table 3). Available data suggest that the proteins of this group are important factors of assembly and maturation of the viral particle and get into the viral cores in association with Gag and GagPol proteins or viral RNA. Some of the factors of vesicular trafficking may be potentially involved in the early stage of HIV-1 infection: the functional genomic screening of factors involved in HIV-1 infection showed that the vesicular RAB6A protein is important for the late phase of reverse transcription in infected cells [58].
The cytoskeletal proteins were also abundant in all core preparations, however, the role of these proteins in HIV-1 virions is still questionable. The actin microfilaments form the plasma membrane cortex, and both actin and microtubular networks are involved in HIV-1 particle assembly [111,112,137], so that the proportion of these proteins in viral particles may be up to 15% of the molar level of Gag [94]. Actin is packaged in the virions probably in association with the NC domain of Gag [111,137,138]. Thus, actin and actin-associated proteins coronin, moesin, filamin, and FLII can get into the viral cores due to the actin interaction with Gag or GagPol. However, involvement of these virion-packaged molecules in post-assembly events of the virus life cycle has not been shown [12]. Our data indicate that the ratio of β tubulin between the viral cores reproduced the concentration ratio of this protein between the producer cells, suggesting the capturing, but not specific incorporation, of this cytoskeletal protein into assembling virions. On the contrary, the cytoskeleton regulation proteins were found to be mostly different in the viral cores from different cell types, which may reflect variability of the profiles of these proteins in virus-producing cells. Some of them, particularly Hsp27, a protein containing a nuclear localization signal [139] and found in the cores of virions from all cell types, can be potentially involved in post-entry steps of infection, although the role of this protein in viral replication remains unknown.
The other category of proteins abundant in HIV-1 viral cores is the molecular chaperones. The profiles of these proteins are very similar in all analyzed samples. Indeed, Hsp27, Hsp40 (DnaJ) co-chaperone, Hsp70, Hsp90, numerous members of TCP1 (Hsp60) tetradecameric complex, as well as peptidylprolyl isomerase cyclophilin A were found in all core preparations (Table 2). Previously, these proteins were identified in purified samples of whole HIV-1 virions and Gag preparations [4,10,23]. Hsp70 was also found in HIV-2, SIV MAC and SIV AGM [11,56]. Since the major function of these proteins is to regulate folding of newly synthesized polypeptides, facilitate intracellular protein transport and assemble multisubunit protein structures [140,141], they likely play an important role in HIV-1 particle assembly, processing and folding of the viral proteins during virion core maturation and maintain structural integrity of the viral core and RTCs [11,33,49]. The early RTC functions, especially organization of reverse transcription, may also depend on the proper activity of incorporated chaperones.
The RNA-binding proteins represent the most diverse group of cellular factors in viral cores. Although we identified 29 RNA-binding proteins in the core preparations, only four of them were found in all core samples ( Table 2). These are (1) regulator of nonsense transcript stability (UPF1), (2) ATP-dependent RNA helicase A (RHA or DHX9), (3) small nuclear ribonucleoprotein component (Snrp116 or EFTUD2), and RNA helicase hBrr2 200 kDa (SNRNP200 or HELIC2). The role of the first two factors in HIV-1 infection was thoroughly characterized before. The UPF1 protein, RNA helicase from the SFI superfamily, involved in translation of Gag polypeptide, was found in virus-producing cells in association with HIV-1 ribonucleoprotein (RNP) along with Pr55Gag, viral RNA and cellular protein Staufen 1 [108]. Thus, UPF1 can be packaged in the virions in association with both Gag and the viral genomic RNA. RNA helicase A (RHA), a member of the DEAD family of proteins which are capable of unwinding the double-stranded RNA structure, was earlier found to be associated with HIV-1 Gag and incorporated into HIV-1 virions in an RNA-dependent manner. Packaging of this protein into HIV-1 virions was important for endogenous reverse transcription [10]. Jeang and Yedavalli suggested that RHA incorporated into HIV-1 viral cores might be important for the reverse transcription in RTCs [95]. A recent study that revealed an important role of this enzyme in the annealing of tRNA Lys3 primer [73], confirmed this suggestion. Two other proteins, SNRNP200 (member of the family of U5 DEXH-box RNA helicases) and Snrp116 (U5 snRNP specific protein, 116 kD), are both members of the U5 group of small nuclear RNA proteins, the spliceosome components, and have not been detected in HIV-1 virions before. Since these proteins are known as important components of splicing machinery required for a spliceosome catalytic activity [142,143], they can be associated with HIV-1 pre-mRNA and remain associated with a mature viral RNA molecule.
Interestingly, our analysis showed higher level of DHX9 (RHA) and SNRNP200 (HELIC2) in the cores of virions assembled in T cells, as compared with the viral cores from the monocyte and MDM models, which did not correlate with the abundance of these proteins in producer cells. Since we did not find significant differences in the RNA and CA protein count between the virions from analyzed cells, observed differences suggest that the mechanism of incorporation of these proteins into the virions (binding to viral RNA or interaction with Gag or/and GagPol) is more effective and likely selective in T lymphocytes, than in monocyte and MDM model cells. Because of importance of RNA helicase A for the reverse transcription in HIV-1 virions and RTCs, we expect that SNRNP200 protein may also be involved in cDNA synthesis or accumulation.
Analysis of earlier published genome-wide screens performed by Warrilow and co-authors [22] to select the host factors potentially implicated in HIV reverse transcription showed that the proteins involved in DNA replication, transcription and repair, as well as proteins of the ubiquitin-proteasome pathway may also be important. Within the 17 DNA-binding proteins detected in our core preparations, only two, ATP-dependent DNA helicase II (XRCC5 or Ku80) and TATA binding protein interacting protein 49 kDa (RUVBL1 or Pontin52), were found in the core of all virions. Two other proteins, minichromosome maintenance complex component 5 (MCM5 or CDC46) and regulator of chromosome condensation protein 2 (RCC2), were identified only in Sup-T1 and activated THP1 cells (Table 2), although Western blot showed presence of MCM5 also in non-activated THP1. The protein RUVBL2 (Reptin52) was not identified by MS/MS in activated THP1, but the fact that in cells this DNA helicase is complexed with the closely related RUVBL1 protein in hetero-dodecamers [144] suggests incorporation of this protein in the virions from all studied cells. Our analysis showed that among DNA binding proteins present in the viral cores from different cell types only XRCC5 and especially MCM5 displayed an increased level of incorporation from the T cells, similar to RNAbinding DHX9 and SNRNP200, whereas core incorporation of RUVBL1 and RUVBL2 reproduced their level in virion-producing Sup-T1 and THP1 cells. The DNA helicase MCM5, a member of the MCM family of chromatin-binding proteins is involved in the initiation of DNA replication and was found to be upregulated during the transition from the G0 to G1/S phase of the cell cycle (RefSeq database). Interaction of this protein with HIV has not been shown before. Another DNA helicase, XRCC5 or Ku80, which is involved in repairing DNA double-strand breaks, was earlier found to be important for viral cDNA circularization, nuclear import and integration [145,146]. However, this function was shown for the protein expressed in the infected target cells, but not for virion-incorporated Ku80. Packaging of both MCM5 and XRCC5 (Ku80) DNA helicases in all viral cores and their high levels in the cores of T lymphocyte-derived virions suggest that the core-incorporated molecules of these proteins can also be involved in processes associated with cDNA processing and/or integration during post-entry steps of infection, especially in T cells.
Within the group of ubiquitin-proteasome pathway associated proteins a total of 9 proteins were detected; three of them were identified in viral cores from all producer cells (Table 2). Earlier, numerous 26 S proteasomeassociated proteins were found in HIV-1 and SIV particles [4,56]. Involvement of the ubiquitin-proteasome system in the budding of lentiviral particles was shown earlier (reviewed in [147]). Since all major domains of the membrane-associated HIV-1 Gag molecules have been shown to be ubiquitinated during virion budding [148], the ubiquitination factors could package into virions in association with Gag and then get into the viral cores. However, the role of virion-associated factors of the ubiquitin-proteasome system in the early steps of HIV-1 infection is unknown.

Conclusions
Taken together, results of our study indicate that the profile of host cell proteins packaged in the cores of HIV-1 virions depends on the type of producer cell. High abundance of certain proteins in the cell increases the probability of their capturing by the virions and hence their presence in the viral cores. However, certain members of functional groups of DNA-and RNA-binding proteins, molecular chaperones, cytoskeletal, vesicular trafficking-associated and ubiquitin-proteasome pathwayassociated proteins were found in the cores of virions from all analyzed cells, suggesting that their incorporation is non-random and that they can be directly or indirectly involved in either the virus assembly/budding or early infection events. Our findings that the abundance of cellular proteins DHX9, (RHA) SNRNP200, MCM5, and XRCC5 (Ku80) within virus-producing cells did not correlate with the abundance seen in cores of produced virions, specifically their unexpected higher packaging in T cells, suggests that the incorporation of these factors in T lymphocytes is more efficient than in myeloid cells. These differences may be associated with variability of localization of these host proteins relative to the sites of virion assembly in different cell types and/or with different localization of virion assembly complexes. The host factors abundant in the viral cores may play a role in subsequent steps of HIV-1 infection, specifically in T cells. Further analysis of the role of these proteins in viral replication might reveal new mechanisms of the modulation of HIV infection by the host proteins and identify new targets for antiretroviral therapeutic interventions.

Cells and viruses
The acute monocytic leukemia cell line THP1 (from S. Tsuchiya) and T lymphoblastoma Sup-T1 cells (from James Hoxie) were provided by the NIH AIDS Research & Reference Reagent Program. The human kidney fibroblasts 293 T/17 was purchased from ATCC (Manassas, VA). All cells were maintained at 37°C and 5% CO 2 in 75 cm 2 tissue culture flasks with RPMI-1640 culture media supplemented with 10% Fetal Bovine Serum, penicillin/streptomycin (100 μg/ml), and L-Glutamine.
The stocks of the HIV-1 virus pseudotyped with amphotropic murine leukemia virus envelope glycoprotein (MLV Env) for synchronized infection of THP1 and Sup-T1 cells were prepared by transfection of 293 T/17 cells with HIV-1 NL4-3 provirus-encoding plasmid [149] and pcDNA-Env(MLV) plasmid (kindly provided by Nathaniel Landau) at a 4:1 ratio using Metafectene transfection reagent (Biontex, Planegg, Germany). The cells were transfected in 75 cm 2 tissue culture flasks; the plasmid DNA containing media was changed at 5 h post-transfection. After overnight incubation with fresh RPMI-1640, the media was changed again and the cells were cultured in fresh media for an additional 48 h at 37°C and 5% CO 2 . Then, the supernatants were harvested, filtered through a 0.45-μm filter and stored on wet ice at 4°C for 1-4 days.

Infection
The viral suspensions were normalized according to their RT activity corresponding to 1 × 10 6 cpm per 1 × 10 6 cells, mixed with Polybrene (Sigma) to a final concentration of 8 μg/ml and used for infection of approximately 200 × 10 6 viable Sup-T1 or THP1 cells by spinoculation [150]. Infection was performed in 6-well plates (5 × 10 6 cells per well) by centrifugation at 1000xg for 2 h at 18°C. After a 2 h incubation at 37°C and 5% CO 2 , the cells were washed from the virus-containing media, re-suspended in RPMI-1640 (pre-warmed to 37°C ) and seeded in a regular (Sup-T1 and non-activated THP1) or polylysine-treated (THP1 for activation) 75 cm 2 tissue culture flasks at a concentration of 4 × 10 6 cells per ml (Sup-T1, non-activated THP1) or of 1 × 10 6 cells per ml (THP1 for activation). To get activated THP1 cells, the PMA and vitamin D 3 solutions were added to cells to a final concentration of 100 nM. Then, the cells were incubated at 37°C and 5% CO 2 for 72 h.

Concentration of virus and "spin-thru" isolation of viral cores
Virus-containing culture media from infected Sup-T1, activated and non-activated THP1 cells, as well as the media from the same types of non-infected cells (control) were harvested at 72 h after incubation with virus (or equivalent volume of the virus-negative culture media) and purified from cell debris by being centrifuged at 2,500 rpm and 4°C for 5 minutes and filtered through 0.45 μm syringe filters. Then, filtered samples were centrifuged at 100,000xg and 4°C for 3 h through 2 ml cushions of 30% sucrose in STE buffer (10 mM Tris-HCl [pH 7.4], 100 mM NaCl, and 1 mM EDTA) in a Beckman SW-41 rotor. The pellets were re-suspended in 300 μl of STE buffer and the viral cores were then isolated by "spin-thru" purification as described earlier [63][64][65][66]. Briefly, 3.8 ml of a 30-50% linear density gradient of sucrose in STE buffer was overlaid with 1 ml of 15% sucrose containing 1% Triton X-100 and then covered with a 0.4-ml cushion of 7.5% sucrose in STE. The HIV-1 positive and negative samples, concentrated through 30% sucrose and resuspended in STE (0.3 ml) were carefully layered on top of the 7.5% sucrose layer and centrifuged in a Type 100 Ti rotor (Beckman Coulter) at 100,000xg and 4°C for 16-18 h. The pellets were re-suspended in 26 μl of STE buffer and replaced to polypropylene, non-siliconized Eppendorf microtubes; 4 μl aliquots were set aside for the p24 CA ELISA assay. The CA p24 Gag -normalized suspensions of HIV-1 cores and control suspensions were subjected to SDS-PAGE protein separation for subsequent LC-MS/MS analysis, Western blotting, or to In-solution protein digestion with trypsin for the LC-MS/MS analysis of unseparated protein samples.
In order to test purity of the "spin-thru" isolated cores from undestroyed viral particles, 400 μl aliquots of the suspensions of viral cores and concentrated whole virions were separately subjected to centrifugation in a 30-70% sucrose gradient for 5 h at 125,000xg and 4°C in a SW-60Ti rotor (Beckman Coulter). Ten fractions of the gradient (each 400 μl) were then collected from the bottom of the tubes and densities were determined. All fractions were dialyzed versus 1 L of ice-cold PBS using Tube-O-DIALYZER 1 kDa MEDI Kit (G Biosciences, St. Louis, MO) according to the manufacturer's protocol and then applied for p24 enzyme-linked immunosorbent assay using Alliance HIV-1 p24 ELISA Kit (PerkinElmer, Waltham, MA).
Additionally, electron microscopy (EM) was applied to test purity of the viral and core preparations. For EM, the virions concentrated through 30% sucrose and "spinthru"-purified core preparations were resuspended in 20 μl of STE buffer, incubated 20 minutes on formware carbon film-coated 100 square mesh nickel grids (Electron Microscopic Sciences) at room temperature, and then incubated with 4% glutaraldehyde fixing solution for 10 minutes. After five-time wash in molecular grade water (Mediatech, Manassas, VA), samples were stained with 2% uranyl acetate. For analysis of virion structure, the pellets of virions after centrifugation through 30% sucrose were washed twice with PBS and then fixed with 4% glutaraldehyde for 4 h. The preparations were further fixed with 2% OsO 4 for 2 h, dehydrated with a graded series of ethanol dilutions ranging from 25% to 100% and then embedded in Araldite 502 resin. Ultrathin sections were contrasted with 2% uranyl acetate in methanol and 1% lead citrate. All preparations were examined on a JEOL JEM 1200 transmission electron microscope operating at 100 kV.

Gel separation of proteins, protein digestion and peptide extraction
The volumes of viral core suspensions, each containing 400 ng of p24 CA protein, and control suspensions taken in twofold excess were mixed with equal volumes of Laemmli Sample Buffer (BioRad, Hercules, CA) containing 5% β mercaptoethanol, heated in boiling water for 2 minutes and applied for SDS-PAGE protein separation. Separation of proteins was performed in 12.5% Tris-HCl Criterion Precast Gel (BioRad) at 100 V and 4°C for 2-2.5 h. The gel was stained in 0.1% (wt/v) Coomassie (BioRad) solution (40% methanol (v/v), 10% acetic acid (v/v) in water with 1 g/L of Brilliant Blue R-250) for 1 h at room temperature. After 7-8 washes in de-staining solution (contains the same components, as staining solution, except Brilliant Blue R-250) the gel was replaced to water, and each lane was sectioned into 10 contiguous pieces, which were subjected to the "in-gel" proteolysis according to the modified previously published protocol [151] Briefly, acetonitrile (ACN) dehydrated gel pieces were rehydrated in 10 mM DTT and incubated at 60°C for 1 h. After cooling at room temperature, the gel slices were incubated with 50 mM iodacetamide for 1 h at room temperature in the dark for alkylation of proteins. After the second dehydration, a 15 μl dose of Trypsin Gold (Promega, Madison, WI) solution (20 μg/ml) in 40 mM NH 4 HCO 3 /10% ACN was added to each of the gel pieces. After 1 h saturation at 4°C, the pieces were incubated at 37°C overnight. The resulted peptides were extracted three times: (1) with 25 mM of NH 4 HCO 3 : ACN (1:1); (2) 5% formic acid (FA); (3) 5% FA:ACN (1:1). After pooling all the extracts together, samples were purified through ZipTip pipette tips C18 (Millipore), eluted with 30 μl of 0.1% trifluoroacetic acid (TFA) in 80% ACN and subjected to HPLC separation and MS/MS analysis.
For "in-solution" protein digestion, the suspensions of HIV-1 cores after "spin-thru" centrifugation were treated with 10 mM DTT (60°C for 1 h) and 150 mM iodacetamide (1 h at room temperature in the dark) in 20 ul of STE buffer. The protein samples were then mixed with 100 μl of 200 mM ammonium bicarbonate and treated with 200 ng of Trypsin Gold (Promega) at 37°C overnight. The resulted peptides were dried in SpeedVac, resuspended in water, purified through ZipTip pipette tips C18 as described above and then subjected to HPLC separation and MS/MS analysis. help with electron microscopy. This work was supported by the District of Columbia Developmental Center for AIDS Research (DC D-CFAR), an NIHfunded program 5P30AI087714-02 and the NIH grant from Research Centers in Minority Institutions (RCMI) Program of the Division of Research Infrastructure, National Center for Research Resources (RCMI-NIH 2G12RR003048). SS is a predoctoral student in the Microbiology and Immunology Program of the Institute for Biomedical Sciences at the George Washington University. This work is from a dissertation to be presented to the above program in partial fulfillment of the requirements for the Ph.D. degree.