' . ‘3 3%‘#¢§5f.' , .4 4‘ n‘ “#535!” 431.35.: .. .p , u. , v. .~ i 4473.3: 1 ,- 9W} .33: $92, . 0 “iii n ‘2‘ ,. fig, L? ‘r‘ , ,5 11,1 5" V w '¢ ‘ 1 1.: Anna: ~15, .‘._ c u ,I - 1. a ‘v 3‘.» V “WhA H‘l. c ms lHIHHllHlillflllllllllIlillllllllllllllllllllllllilllllll 3 1293 01399 2239 DJ This is to certify that the dissertation entitled Ectopic Expression of Chicken HMG14A and HMG17 Chromatin Binding Proteins presented by Natalie Brown has been accepted towards fulfillment of the requirements for PH.D. degree in microhiology MS U is an Affirmative Action/Equal Opportunity Institution 0— 12771 ECTOPIC EXPRESSION OF CHICKEN EMGliA AND EMGl7 CERGMATIN BINDING PROTEINS BY Natalie S. Brown A DISSERTATION Submitted to? Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Microbiology 1995 ABSTRACT ECTOPIC EXPRESSION OF CHICKEN HMG14a AND HMG17 CHROMATIN BINDING PROTEINS BY Natalie S. Brown The high mobility group (HMG) chromatin binding proteins HMG14 and HMG17 are abundant and highly conserved eukaryotic proteins. Although their function remains unknown,' these proteins are knOwn to bind to nucleosomes preferentially in regions of actively transcribing chromatin. To better understand these proteins and the role they play within the cell, we have ectopically expressed wild type and mutant chicken HMG14a and HMG17 cDNAs in QT6 (quail fibroblast cell line) cells. Wild type and mmtant cDNAs were introduced into several eukaryotic expression vectors and transfected, transiently and stably, into the QT6 cell line. Stable expression of the wild type HMG17 cDNA resulted.int3 to 5 fold increases in.the level of detectable protein. The overproduction of this protein did not effect the cell viability, growth rate, or cause any gross morphological changes. Chromatin structural studies of the cell lines overexpressing HMG17 demonstrated moderately increased binding of HMG17 to individual nucleosomes yet there was no effect seen on the gross chromatin structure. The production of wild type and mutant HMG14a fusion proteins containing an immunologically distinct C-terminus allowed us to distinguish between the endogenous and exogenously expressed proteins. Using a tetracycline regulated expression system designed to increase the amount of exogenous protein expressed in cells, we transiently expressed mutant, wild type, and fusion HMG14a clones in QT6 cells. We demonstrated that the cells can express these constructs with no difference seen in transfection efficiency, growth characteristics, or level of expression between the mutants and the wild type. Using immunofluorescent confocal microscopy, we examined the in-situ location of exogenous wild type and mutant HMG14a fusion proteins in our transfected cell lines. We demonstrated that these proteins do localize exclusively to the nucleus. The staining patterns seen in these cell lines suggest the HMG14a proteins are in fact binding to the dispersed chromatin of the interphase cell as well as the condensed chromosomes in dividing cells. There was no detectable difference seen between the wild type and mutant proteins used in these studies. Dedications To Bill and Manitou ACKNOWLEDGMENTS I am extremely grateful to my advisor, Dr. Jerry Dodgson for his advice, help, support and patience during my graduate work. I feel it a privilege to have worked with him and know I have benefitted greatly from having him as a role model. I am especially grateful to the members of my committee, Drs. Richard Schwartz, Michelle Fluck, Larry Snyder and Birgit Zipser for their guidance and encouragement throughout the years. I would like to thank the many people who have helped me with their friendship and expertise including my fellow labmates, Huei Min Lin, Yi Li, Kyoung Eun Kim, Don Salter, Bill Payne, Lyman Crittenden, and Wynne Lewis, Dr. Lee Velicer, who always had an encouraging word, Dr. Henry Hunt, who greatly helped me with research advice, and my fellow graduate students, especially Larry Martin, Nicco Yu, Sue Dahger, Michelle Anderson, and Brenda Kenney. I feel it was a great privilege to work among the excellent faculty and support staff of the Department of Microbiology and wish them all continued success. ii TABLE OF CONTENTS List of Figures ........................................... iv Introduction .............................................. 1 Chapter I: Literature Review: The HMG Chromosomal proteins and chromatin struture ................. 3 Chapter II: Materials and Methods .......................... 54 Chapter III: Design and production of chicken HMG14a and HMG17 mutants ............................. 76 Chapter IV: Expression of chicken HMG14a and HMG17 proteins in tissue culture cells ............... 105 Chapter V: Expression of HMG14aFLAG in tissue culture cells ................................... 152 Chapter VI: Chromatin structural studies of cell lines ectopically expressing chicken HMG14a and HMGl7 .......................................... 182 Chapter VII: Immunofluorescence of cells expressing HMG14FLAG ..................................... 195 Chapter VIII: Summary ...................................... 211 List of references. ........................................ 217 Appendix I ................................................. 223 Appendix II ................................................ 224 iii LIST OF FIGURES Chapter I Figure 1 - Progressing orders of chromatin condensation .................................... 6 Figure 2 - Digestion of chromatin with endonuclease .................................... 9 Figure 3 - Alignment of HMG14 and HMG17 protein sequences ...................................... 22 Figure 4 - A: Translated coding region 00 chicken HMG14a cDNA ......................... 25 B: Translated coding region of Chicken HMGl7 cDNA .......................... 27 Figure 5 - Binding of HMG14/17 on nucleosome core ........................................... 31 Figure 6 - Dot matrix comparison of HMG14 and HMG17 cDNA sequences ........................... 43 Figure 7 - Nucleotide sequence of chicken HMG14a cDNA .................................... 45 Figure 8 - Nucleotide sequence of chicken HMG17 cDNA ..................................... 47 Chapter II Figure 1 - TFANEO subcloning strategy ..................... 57 Chapter III Figure 1 - DNA and amino acid sequence of chicken HMG14a and HMG17 DNA binding domain ............ 81 Figure 2 - Mutations produced in the HMG14a and HMGl? cDNAs .................................... 83 Figure 3 - Oligonucleotides used for site-specific mutagenesis in HMG14a and HMG17 cDNAs .......... 85 Figure 4 - in-vi tro mutagensis reaction scheme ............ 89 Figure 5 - Oligonucleotides used for sequencing mutants in the HMGl4a and HMGI7 cDNAs .......... 91 Figure 6 - PCR mutagensis reaction scheme ................. 94 Figure 7 -.A: DNA sequence and location of mutants in the HMG14a cDNA .................. 97 B: DNA sequence and location of mutants in the HMG17 cDNA ................... 99 iv Chapter IV Figure 1 - Map of expression vector TFANEO ............... 110 Figure 2 - Northern analysis of total RNA isolated from QT6 cells transfected withHMG14acDNAconstructs ................... 112 Figure 3 - Northern analysis of total RNA isolated from QT6 cells transfected with HMG17 cDNA constructs .................... 115 Figure 4 - Western blot of QT6 transfected cell lines .................................... 119 Figure 5 - Western blot of QT6 transfected cell lines .................................... 123 Figure 6 — Western blot of QT6 transfected cell lines .................................... 125 Figure 7 - Western blot of QT6 transfected cell lines .................................... 127 Figure 8 - Growth curve for control and transfected QT6 cells ......................... 130 Figure 9 - Map of vectors used in tetracycline regulatable expression system ................. 136 Figure 10 - Western analysis of protein isolated from cell lines transiently transfected with the coding region of HMG14a ............. 140 Figure 11 - Western analysis of protein isolated from cell lines stably transfected with the coding region of HMG14a ............. 143 Figure 12 — Western analysis of protein isolated from cell lines stably transfected with the coding region of HMG14a ............. 145 Figure 13 - Western analysis and luciferase assay of protein isolated from stably transfected clones transiently transfected with pUHD10-3 containing the luciferase gene .......................... 148 Chapter V Figure 1 - Oligonucleotide primers for PCR production of HMG14 FLAG constructs ............ 156 Figure 2 - Schematic of HMG14aFLAG construct and comparison of C-terminus of HMG14a with and without FLAG sequences ............... 158 Figure 3 - Detection of HMGl4aFLAG by Western blot .......................................... 161 Figure 4 - Detection of HMGl4a and HMG14aFLAG in cell fractions from transiently transfected cells ......................................... 164 Figure 5 - Western blot of protein isolated from stably transfected cell lines ............ 168 Figure 6 - Western blot of protein isolated from stably transfected cell lines ............ 170 V Figure 7 - PCR strategy for detecting HMGl4aFLAG/ pUHD10-3 sequences in stably transfected cell lines ........................ 174 Figure 8 - Western blots of protein isolated from stably transfected cell lines ............ 177 Chapter VI Figure 1 - Micrococcal nuclease digestion of chromatin ..................................... 188 Figure 2 - Analysis of nucleoprotein complexes ........... 191 Chapter‘VII Figure 1 - Immunofluorescent staining of cells transfected with HMGl4aFLAG ................... 199 Figure 2 - Negative transfection control ................. 202 Figure 3 - Immunofluorescent staining of cells transfected with HMG14aFLAG ................... 204 Figure 4 - Imunofluorescent staining of cells transfected with HMG14C3FLAG .................. 207 APPENDIX II Figure 1 — Southern analysis of stable cell lines transfected with H3 . 3B and H3 . 2 cDNAs ......... 230 Figure 2 - Northern analysis of stable cell lines transfected with H3 . 3B and H3 . 2 cDNAs ......... 232 vi INTRODUCTION The high mobility group (HMG) chromatin binding proteins HMG14 and HMG17 are abundant and highly conserved eukaryotic nuclear proteins. Their high level of evolutionary conservation and wide distribution suggest they play an important role in chromatin structure. Although the exact nature of their cellular function remainS'uncertain, these proteins are known to bind.to nucleosomes with specific stoichiometery. A variety ofin-vitro experiments suggest that HMG14 and HMG17 are preferentially associated with regions of active or potentially active chromatin. The apparent exclusive association of these HMG proteins with transcribable sequences suggests that they may be involved in some aspect of transcriptional regulation, yet they have been found not to function directly as transcriptional activators. Much of the investigation of these proteins has focused on their physical structure and direct interactions with nucleosomes. With the recent availability of cloned genes for the chicken HMG14 and HMG17 and suitable expression vectors, we chose to study the expression of these proteins in a tissue culture system. In this thesis, we demonstrate stable and transient expression of wild type, mutant, and fusion chicken HMG14a and HMG17 in a 1 quail fibroblast cell line. In stable transfections we have elevated the level of HMG17 to 3-5 times that of the normal cell. This was achieved with little or no effect on the cell phenotype, growth characteristics, or chromatin structure, suggesting these cells can tolerate at least this much excess protein” We have also shown that these proteins localize to the nucleus with in-situ immunofluorescent studies and appear to bind to the dispersed chromatin of the interphase cell. Chapter 1 Literature Review: The HMG chromosomal proteins and chromatin structure 4 Chromatin Structure and Active Genes The complex of DNA and protein in the nucleus of eukaryotic cells is termed chromatin. The dynamic structure of chromatin is evident in the mitotic processes of chromosome condensation and relaxation and the cellular mechanistics of DNA replication and transcription. The basic repeating unit of chromatin is the nucleosome. The core nucleosome is composed of a 146 base pair duplex DNA wrapped twice around around an octamer of histone proteins. This octamer is composed of two each of the four core histones, H2A, H2B, H3, and H4. The nucleosomes are linked together by the DNA duplex and this linker DNA.varies in length, producing the heterogeneity seen among nucleosomes from.different cells and at different times in the developmental process. A fifth histone, H1, is associated with this linker DNA and is known to bind to the DNA strand as it enters and exits the core nucleosome. The nucleosome is also involved in the higher orders of structure that compress the genomichNA.into the compact form found in the nucleus. The nucleosomal fiber, which appears as "beads-on-a-string" under an electron microscope and has a diameter of 10 nm is coiled into a 30 nm fiber helped by the addition of histone H1 which acts as a crosslinker between nucleosomes. This 30 nm fiber is then subject to further levels of condensation depending on the mitotic state of the cell, among other things, which is brought about by the 5 binding of specific nonhistone proteins which form the scaffold of the chromosome. An illustration of the various levels of chromatin condensation is shown in figure 1. The condensed structure of chromatin is neccessary to compact the large amount of DNA.present in the nucleus. The human genome of 3 x 109 base pairs would extend over one meter if completely extended, yet is packaged into the nucleus with an average diameter of 10 um. The dynamic state of chromatin is evidenced by the process of gene transcription. Highly condensed chromatin is too compact to allow the cellular transcriptional machinery to access the DNA. Therefore, the regions of DNA which are actively being transcribed or are in a potentially active state, approximated to be 10 - 20 % of the total genome, must be in a more accessible or less compact structure. This decondensed structure of active genes can be confirmed experimentally with the use of non-specific endonucleases such as DNaseI. It has been shown that DNaseI will preferentially digest active or potentially active gene sequences and that this increased sensitivity is a result of changes in chromatin and nucleosomal structure specific to these regions in the chromatin (1). These changes in the structural organization of active genes can also be seen by digesting chromatin with micrococcal nuclease which produces characteristic patterns of digestion that distinguishs bulk chromatin from the coding regions of Figure 1 - Progressing orders of chromatin condensation An illustration of the progressing orders of chromatin condensation giving rise to the highly condensed metaphase chromosome (83). Figure 1 8 active genes (2). An illustration of the digestion of chromatin with DNaseI is shown in figure 2. The changes in chromatin structure associated with active genes are thus illustrated by the increased susceptability of these regions to nucleases and are due to a decondensing of the chromatin at these locations. This allows the nuclease molecules increased physical access to the DNA. These changes are coincident with modification of the histone proteins, possible substitution of the histones with variants, loss of histone H1, and the presence of nonhistone chromatin binding proteins such as the high mobility group (HMG) proteins on the nucleosomes. Analysis of the core histone proteins found in chromatin fractions enriched for active sequences shows these proteins to be highly acetylated on lysines near their amino terminus. Although this modification produces subtle changes in the nucleosome conformation (5), it does not appear to be sufficient to generate the active chromatin conformation or to increase the rate of transcription (3,4). Acetylation at the histone tails may reduce the helical periodicity of the linker DNA and by loosening interactions at the periphery of the nucleosome, help to dissociate the histone H1, which is found in reduced amounts in active chromatin (6). The two small, basic chromatin binding proteins, HMGl4 and HMG17, are the most abundant nonhistone chromatin binding proteins found in the nucleus of higher eukaryotes and have Figure 2 - Digestion of chromatin with endonuclease An illustration of the digestion of chromatin with pancreatic DNaseI (83). Endonucleases such as DNase I cut chromatin first at nuclease hypersensitive sites characteristic to a particular gene. This is followed by selective degredation of the DNA sequence of actively transcribing and potentially active genes. The distinct "open" conformation in these regions of chromatin allows for this selective digestion. HMG14/17 are selectively associated with these regions of active and potentially active chromatin. _1o Figure 2 11 been directly associated with active chromatin. Weisbrod et al showed that the DNaseI sensitivity of active gene sequences is lost upon extraction of the chromatin with 0.35M NaCl, which releases a fraction containing the HMG14 and HMG17 proteins. Upon reconstition with the 0.35 M NaCl fraction or with a subfraction highly enriched for HMG14 and HMG17, this sensitivity is restored (7,8,9). Others have shown that mononucleosome fractions enriched for active sequences are also enriched for HMG14 and HMG17 (12). Numerous other lines of evidence support the association of these HMG proteins with regions of active chromatin. HMG14 and HMG17 proteins immobilized on agarose columns can selectively bind to and retain actively transcribed nucleosomes (10) . HMG17-specific antibodies have been used to selectively isolate oligonucleosomes from active chromatin (11), and microinjection of antibodies to HMG17 in somatic cells has been shown to inhibit transcription (13). Although the correlation.between active gene structures and the HMGl4 and HMG17 proteins appears well established, the actual function of these proteins and the role they play in the generation and/or maintanance of active chromatin remains unknown. 12 The HMG chromosomal proteins The high mobility group (HMG) chromatin binding proteins are the most abundant and ubiquitous nonhistone proteins found in the nucleus of eukaryotic cells. These proteins are about one tenth as abundant as the histone proteins at approximately 10° molecules per cell. The HMG proteins have been highly conserved throughout evolution suggesting an important role in cellular function. Despite extensive investigation, their roles remain virtually unknown. The HMG proteins were first identified by Johns and co -workers in their attempts to isolate histone H1 from calf thymus chromatin. In addition to histone H1, the nuclear preps contained a number of additional chromatin proteins with distinct charge and size properties. It was found that these nonhistone proteins could be removed from.chromatin with 0.35 M NaCl extraction and further purified as a 2% TCA (trichloroacetic acid) soluble fraction (14). When this 0.35 M NaCl extractable, 2% TCA soluble fraction was run on polyacrylamide gels, up to 16 bands were identified, all of which ran with a high mobility due to their small size and highly charged nature. This group of proteins was thus named the high mobility group proteins and designated HMGI, HMG2 and so on according to their position in.the gels. Later, it was determined that most of these bands were 13 proteolytic degradation products. HMGl, HMG2, HMG14, and HMGI7 are now recognized as the major HMG proteins in vertebrate tissue, with HMGI and HMGY found in at least some species. As other groups continued to find these proteins in other organisms, it became apparent that they constituted a family of abundant nuclear proteins with distinct properties and a probable important role in cellular function. The working definition of the HMG family of proteins was initially derived from the properties of the calf thymus proteins, these properties for the HMG proteins are: (14) 1. chromatin proteins 2. extractable from chromatin in 0.35 M NaCl 3. not precipitated from the above extract in 2% TCA 4. high in basic amino acids, approximately 25% or more 5. high in acidic amino acids, approximately 20-30% 6. relatively high in proline, 7% or more 7. soluble in 5% perchloric acid This definition is somewhat problematic today as proteins isolated from lower eukaryotes, which are probably functionally related, may be excluded.based upon differences in one or more of these properties. Isolation and characterization of the cDNAs for the HMG proteins from a variety of organisms has allowed this defintion to be re -evaluated and now HMG proteins are classified based primarily upon sequence similarities as well as their 14 physical properties. HMG-like proteins from lower eukaryotes HMG-like proteins have been isolated from a number of lower eukaryotic species (15). These include Drosophila melanogaster, ceratitis capitata, Saccharomyces cerevisiae, Neurospora crassa, Dictyostelium discoideum, Physarum ,polycephalum, Aspergillus nidulans, Tetrahymena pyriformis, and TEtrahymena thermophila. The HMG-like proteins isolated from these organisms have physical characteristics similar to the HMGs of vertibrates such as, solubility in 5% perchloric acid, gel mobility and amino acid composition but few have been analyzed.furthert Protein sequence determination of the HMG-like proteins NHP6a, NHP6b and ACP2 found in saccharomyces cerevisiae show that they are homologues to HMGI, yet lack some characteristic internal repeats found in these proteins. Deletion of the ACP2 locus from S. cerevisiae has been shown to be haploid lethal (15). The LGI protein from T. thermophila has also been shown to have partial homolgy to HMGI and is thought to share a common ancestor with the mammilian HMGI protein. Little else is known of these proteins and their function remains obscure. The HMGI/r family The HMGI family is composed of HMGI and it's isoform.HMGY. 15 These proteins are not detectable in terminally differentiated, non -proliferating cells but are enriched in rapidly growing, transformed or malignant tissues (15) . These proteins resemble the HMG14 and HMG17 proteins and although very little is known of their function, it is thought they may replace HMG14 and.HMG17 and influence cell growth and/or gene regulation. The full length cDNAs have been isolated for human HMGY and HMGI and for HMGY in the mouse (15). The two proteins are isoforms resulting from.differential processing of the RNA. The HMGI and HMGY mRNAs are expressed preferentially in malignant, non-differentiated and neoplastic tissues. Induction of differentiation in murine teratocarcinoma cells results in down-regulation of mRNA expression (15). Thus it appears that the level of HMGI and HMGY expression is coupled to the proliferative state of the cell, yet it is not known whether this regulation is a consequence of or a prerequisite for differentiation. The HMGI and HMG2 family The HMG1 and.HMG2 proteins constitute the most abundant HMG family, at approximately one molecule per 3000 base pairs of DNA. HMG1 and HMG2 have molecular weights of 27 and 28 Kd respectivelyx These proteins can bind both single and double stranded DNA yet show a preference for binding to single stranded DNA (15). In addition, they show a preference for 16 binding cruciform DNA in-vitro though the presence of this form of DNA in-vivo is debatable. Experimental results on their cellular role are somewhat conflicting and their function remains unknown. Several lines of evidence suggest they may play a role in chromosomal replication (15). Antibodies to HMG1 and I-IIVIGZ microinj ected into nuclei inhibit DNA synthesis and the proteins can stimulate DNA polymerases (15). Antibodies to HMG1 were shown not to inhibit RNA synthesis by RNA.polymerase II in somatic cells, yet microinjection into amphibian oocyte nuclei showed retraction of the transcription loops, suggesting that these proteins stimulate transcription (17). These proteins have also been shown to increase the rate of binding of some transcription factors to their DNA recognition sequences (15). In contrast to HMG14 and HMG17, these proteins do not bind selectively to regions of-active chromatin” Recently, it has been observed that HMG1 and HMG2 can.bend.DNA.efficiently into small circles and may facilitate cooperative interactions between cis-acting factors by promoting DNA flexibility. The ability to promote highly compact forms of DNA may indicate a general role for these HMG proteins in chromatin structure (19). HMGl and HMG2 proteins have been isolated and sequenced from.a variety of sources including rat, bovine, human, pig, and Chinese hamster ovary cells. Alignment of the various 17 HMG1 protein sequences show them to be 97-99% homologous to the human.proteins. HMG1 is 214 amino acids in length and can be divided into three domains (15). The first two N-terminal regions, up to amino acid 164, constitute an internal repeat separated by 10 amino acids from residue 80-89. These domains are thought to have a globular structure and are rich in charged amino acids with a net positive charge of +20. This region is most likely involved in binding to the DNA. The C -terminal portion of the protein is unstructured with the last 29 residues being highly acidic (15). This region may be involved with binding to positively charged chromatin proteins such as the histones. The distibution of charges along the HMG1 molecule is thus highly asymmetric. A number of unrelated eukaryotic DNA.binding proteins contain regions homologous to the N-terminal repeats found in HMG1. This motif has been named the HMG box and is involved in the DNA binding ability of these various proteins. cDNA clones have been isolated for HMG1 from several tissues from various species including human, bovine, rat, pig and Chinese hamster ovary cells (CHO). No cDNAs for HMG2 have yet been isolated. All mammalian cells contain three mRNA species for HMG1 as detected by northern analysis. The sizes of these transcripts are 2.4 kb, 1.4 kb and 1 kb, with the 2.4 kb band being the predominant species (15). This heterogeneity may be due to differential processing of a 18 primary transcript. One of these species may be the transcript for HMG2 as the primary sequence of the proteins predicts similarities in the transcript sequences. The various cDNAs for HMG1 are highly conserved showing a 93% sequence identity among them.in the open reading frame (15). Portions of the 3' untranslated sequence show very high levels of homology with the first 300 bp homologous at about the 90% level (15). This suggests that this region may be important for message processing. Southern analysis suggests there may be multiple copies of HMG1 and HMG2 genes in the genome of the cells and tissues examined. It is probable some of these represent psuedogenes. No genomic clone has yet been isolated for HMG1 or HMG2. The level and cellular distribution of HMGl and HMGZ proteins have been investigated in mammilian tissues (17). These proteins are present in the nucleus and the cytoplasm and the subcellular distribution as well as the amount of protein are highly tissue specific. In general, in slow growing or differentiated cells, HMG1 and HMG2 are found predominantly in the cytoplasm. In actively dividing cells, these proteins are found in the nucleus. In addition, the levels of HMGI and HMGZ are inversely correlated with the level of histone Hr’(17). These results are consistant with a role of HMG1 and.HMGZ in replication or transcription. 19 The HMG14 and HMGI? family HMG 14 and HMG17 are the smallest HMG proteins with molecular weights of 14 and 11 kd. They are the only known DNA binding proteins with a higher affinity for nucleosomes than naked DNA. The HMG14 and HMG17 family of proteins are the best characterized HMGs and, although their physical properties and specific interactions with the nucleosome are well understood, the function of these proteins remains controversial. Characteristics of the HHGI4 and HIGI7 proteins HMGl4 and HMG17 proteins have been isolated and analyzed from a variety of species including human, calf, mouse, and chicken. Trout chromatin protein H6 is closely related to HMG14 and HMG17 and is considered a member of this family. Analysis of the protein sequences of these HMG proteins shows them to be highly conserved with stretches of amino acids remaining invariant in all species tested. The HMG17 proteins appear to have evolved slower than HMG14 since their sequence is more highly conserved. The sequence homolgy among the various HMG17 proteins is over 91%, while the sequence homolgy among the HMG14 proteins ranges from 49-94% (15). HMG14 and HMG17 share several highly conserved elements and have over 30% of their amino acid sequence in common; yet they are quite distinct protein families, the 20 similarity between any members of the two groups being less than 52% (15). Trout H6 is approximately 65% similar to the HMGl7 proteins and 55% similar to the HMG14 proteins (15). The major chicken HMG14 protein, called HMG14a, is also intermediate in similarity between the the HMG14 and HMG17 subfamilies. The high level of conservation seen among the HMG proteins, representing an evolutionary span of more than 400 million years from trout to human, suggests probable functional significance for those regions showing the highest evolutionary constraints. The chicken HMG14 and HMGI7 proteins are 104 and 89 amino acids in length, respectively. They can be divided into several domains based upon sequence homology, charge and function. The N-terminal regions of the molecules are basic and contain the DNA binding domains which have a high positive charge. The acidic C-termini have a negative net charge and may mediate interactions with positively charged proteins such as the histones. The C-termini of two HMG molecules bound to the same nucleosome may also interact with each other to stabilize the particle and/or facilitate a cooperative mode of binding. The homology of the proteins vary along the polypeptide chain. The first four amino acids of HMG14 and HMG17 are PKRK (appendix I) and are absolutely conserved among species. The location of the DNA binding domain varies slightly between species. In the 21 chicken proteins, the HMG14a DNA binding domain is located from residue 16-40 and the HMG17 from 21-45. In the calf proteins, DNA binding has been localized to amino acids 17 -60 in HMGl4 and 15-40 in HMG17 (21). Regardless of its location, the DNA binding domains of these proteins contain two stretches of virtually invariant residues, P(K,Q)RRSARLSA and KPKKA. An alignment of the known protein sequences of HMG14 and HMG17 is shown in figure 3. Analysis of the primary protein structures of HMGI4 and HMGl7 also shows conservation in the distribution of charged residues along the polypeptide chains. The first 20 amino acids of HMG17 contain 12 charged residues with a net charge of +2. Similarly, the first 15 amino acids of HMG14 contain 7-8 charged residues with a net charge of +1. The central region of these molecules, which contains the DNA binding domain, is highly positively charged with residues 20-64 in HMG17 and 15-79 in HMG14 having net positive charges of +16 and +15 respectively. The DNA binding domains in this central region, approximately 25 amino acids long, have basic/acidic ratios of 9/1. The C-terminal 26 amino acids of HMG17 contains 9 charged residues with a net charge of -3. In HMG14, the C-terminal 23 amino acids contain 18 charged residues with an overall charge of —3. This high number of charged residues in the C—terminus is characteristic of the other HMG protein families as well. 22 Figure 3 - Alignment of HMG14 and HIGI7 protein sequences Alignment of HMG14 and HMG17 protein sequences from various organisms (15). Overlined sequences indicate regions of invariant amino acids or those which are conservatively substituted among the proteins. 23 human HMG14 PKRKVSSAEG -------- AKEE~PKRRSARLSAKPPAKVEAKPKKK calf HMG14 PKRKVSSAEG ------- AAKEE-PKRRSARLSAKPAPAKVETKPKK mouse HMG14 PKRKVS-ADG ------- AAKAE-PKRRSARLSAKPAPAKVDAKPKK chicken HMG14a PKRKAP ~AEGE ------- AKEE - PKRRSARLSAKPAPPKPEPKPKK chicken HMG14b PKRKVAASRG -------- GREEVPKRRSARLSAKPVPDKAEPKAKA human HMG17 PKRK---AEGDAKGDKAKVKDE-PQRRSARLSAKPAPPKPEPKPKK calf HMG17 PKRK---AEGDAKGDKAKVKDE-PQRRSARLSAKPAPPKPEPKPKK mouse HMG17 PKRK---AEGDAKGDKTKVKDE-PQRRSARLSAKPAPPKPEPKPKK ChickenIHKIU7 PKRK---AEGDTKGDKAKVKDE-PQRRSARLSAKPAPPKPEPKPKK trout H5 PKRKSA ----- TKG ------ DE-PARRSARLSARPVP-KPAAKPKK AAAK ------ DKSSDKKVQTKGKRGAKGKQ-AEVANQETKEDLPAENG AAGK ------ DKSSDKKVQTKGKRGAKGKQ-AEVANQETKEDLPAENG AAGK ------ DKASDKKVQIKGKRGAKGKQ-ADVADQQTT-ELPAENG AAPKKEKAANDKKEDKKAATKGKKGAKGKD--ETKQEDAKEENHSENG LAAK ------ DKSENKKAQSKGKKGPKGKQTEETNQEQIKDNLPAENG APAK ------- KGE--KVP-KGKKG -------- KADAGKEGNNPAENG APAK ------- KGE--KVP-KGKKG -------- KADAGKDGNNPAENG APAK ------- KGE-—KVP-KGKKG -------- KADAGKDANNPAENG AAPK ------- KSE--KVP-KGKKG -------- KADAGKEGNNPAENG AAAP ----------- KKAV-KGKKA ------------------- AENG ETKTEESPASDEAGEK-EAKSD ETKNEESPASDEAEEK-EAKSD ETENQ-SPASEE--EK-EAKSD DTKTNEAPAAEASDDK-EAKSE ETKSEETPASDAAVEKEEVKSE DAKTDQAQK--AEGAG-DAK-- DAKTNQAEK--AEGAG-DAK-- DAKTDQAQK--AEGAG-DAK-- DAKTNQAEK--AEGAG-DAK-- DAKAEAKVQAAGDGAG-NAK-- Figure 3 24 The specific amino acid contents of the chicken HMG14a and HMG17 proteins is listed in figure 4. Interestingly, all HMG14 and HMG17 proteins have an unusually high number of proline residues. In chicken HMG17, there are 8 proline in a stretch of 17 residues in the DNA binding domain. Likewise, in chicken HMG14a, there are 7 prolines out of 17 residues. Proline residues will produce a bend in a polypeptide chain, normally disrupting or preventing regions of secondary and tertiary structure from forming in the molecule. The high level of proline in these HMG proteins most likely contributes to the lack of higher order structure seen in these polypeptides. Using circular dichroism and NMR studies, it was shown that these proteins do not form any secondary or tertiary structure over a wide range of salt conditions (21,22). This suggests that the conformation of these proteins is controlled by the binding to other factors in the chromatin, particulary DNA. However, analysis of the HMG14 and HMG17 protein sequences by the Chou-Fasman secondary structure prediction model suggests the acidic C-termini could be in a weak alpha helical conformation with the acidic residues positioned on a single face of the helix (21,22). This lends support to the suggestion that this region participates in protein-protein interactions primarily through the acidic face of the helix as is seen 25 Figure 4A - Translated coding region of chicken HMGl4a cDNA The translated coding region of chicken HMG14a is shown. The total number and percent content of each amino acid is listed. The predicted molecular weight is 11,225 daltons. See appendix I for amino acid abbreviation list. CCC AAA AGA AAG GCT CCA GCT GAA GGC GAG GCG AAG GAG GAG CCA P K R K A P A. E G E A K E E P 15 AAG AGA.AGG TCG GCC AGA CTA TCT GCT AAA CCT GCT CCG CCT AAA K R R S A R L S A K P A P P K 30 CCG GAG CCA AAG CCC AAA AAG GCA GCA CCT AAG AAA GAA AAG GCA P E P K P K K A A P K K E K A 45 GCA AAC GAT AAA AAG GAA GAC AAA AAG GCA GCA ACA AAA GGG AAG A N D K K E D K K A A T K G K 60 AAA GGA GCC AAA GGC AAA GAC GAA ACT AAA CAA GAG GAT GCA AAA K G A K G K D E T K Q E D A. K 75 GAA GAA AAC CAC TCT GAA AAT GGA GAT ACC AAA ACT AAT GAG GCA E E N H S E N G D T K T N E A 90 CCA GCT GCT GAA GCA TCT GAT GAT AAG GAA GCC AAG TCC GAG P A A E A S D D K E A K S E 104 26 Total number Total percent Amino acid 1072080080050888000 8065041041001343000 l l 1 9076051061011454000 1 l 2 l ACDEFGHIKLMPQRSTVWY Figure 4A (con.) 27 Figure 4B - Translated coding region of chicken HMG17 cDNA The translated coding region of chicken HMG17 is shown. The total number and percent content of each amino acid is listed. See appendix I for amino acid abbreviation list CCG AAG AGA AAG GCT GAA GGA GAT ACC AAG GGC GAT AAG GCC AAA P K R K A E G D T K G D K A K 15 GTT AAG GAT GAG CCA CAA CGG AGA TCG GCA AGG TTA TCT GCT AAA V K D E P Q R R S A R L S A K 30 CCT GCC CCT CCG AAG CCA GAG CCT AAA CCT AAA AAG GCA GCT CCA P A P P K P E P K P K K .A A P 45 AAG AAG AGT GAG AAG GTG CCC AAG GGA AAG AAG GGG AAA GCT GAT K K S E K V P K G K K G K A D 60 GCT GGC AAG GAG GGA AAC AAC CCT GCA GAA AAT GGA GAT GCC AAA A G K E G N N P A E N G D A K 75 ACA GAC CAG GCA CAG AAA GCC GAA GGT GCT GGT GAT GCC AAG T D Q A Q K A E G A G D A K 89 28 Total number Total percent Amino Acid 90990100710444542200 60770000410323432200 1 1 2 1 50770900210313432200 1 2 1 ACDEFGHIKLMNPQRSTVWY Figure 4B (con.) 29 with other classes of proteins such as transcription factors. Association of HMGI4 and HMG17 with active chromatin HMG14 and HMG17 are the only DNA binding proteins with a higher affinity for nucleosomal DNA than naked DNA. In early experiments to determine the stoichiometry of HMG binding to the nucleosome, nucleosomal core particles were salt stripped with 0.35M NaCl to release the HMG proteins, then titrated with HMG14 or HMG17-and electrophoresed on native gels (22,23,24). Gel shift patterns clearly show two more slowly migrating bands indicating two potential binding sites per nucleosome for HMG14 or HMG17. These binding sites were located on the inner face of the DNA as it emerges from the nucleosome core. There was no difference seen when using HMG14 or HMG17 alone or in combination in these reconstitution experiments. This suggests that these two proteins bind with equal specificity to the nucleosomal core particle. Recently, it was shown that a 30 amino acid long peptide corresponding to the DNA binding region of HMG17 can bind with equal affinity and specificity to that of the intact protein (32). Recent analysis of the specific binding of the HMG14 and HMG17 to nucleosome cores and H1 depleted chromatosomes confirmed the stoichiometry of two nucleosomes per particle 30 and found the path of HMG14 on the surface of the nucleosome is indistingiushable from that of HMG17 (26). Using hydroxy radical cleavage mapping experiments, it was shown that these HMG proteins protect DNA 25 base pairs out from the core particle on either end (26). The sites occupied by HMG14 and HMG17 on the ends of the core particle are distinct from the sites occupied by the linker histones (26). A model of the binding of HMG14 and HMG17 to the nucleosome is shown in figure 5. Other studies of the specific binding of these HMG proteins to the nucleosome further characterize this interaction. Using circular dichroism and thermal denaturation, it was shown that the binding of HMG17 produces an overall stabilization and condensation of the core particle (26). Neutron scattering experiments demonstrated that the binding of HMG14 produced a slight increase in the radius of gyration of the DNA and that there are fewer nucleosomes per repeat in HMG14 containing fibers (27,28). Binding of HMG14 or HMG17 to the nucleosome core does not induce any detectable rearrangment of the core histones but may promote a slightly more compact core histone structure (27). Crosslinking studies show the C—terminus of HMG17 interacts primarily with the core histone H2A (29,30). Although every nucleosome has two potential binding sites for HMG14 and HMG17, the limited amount of these proteins in 31 Figure 5 - Binding of HMG14/17 on nucleosome core Location of HMG14 and HMG17 on the nucleosome core (25). The model illustrates binding of the HMG proteins to the nucleosome core. The nucleosome core potentially contains two HMG molecules which bind to the DNA as it enters and exits the core structure. The acidic tails of the HMG molecules are free to interact with core histones or each other. 32 Figure 5 33 the nucleus restricts this binding to a subset. It is estimated that 1 out of every 10 nucleosomes will have HMG bound in vivo. This coincides with the estimated 10-20% of the chromatin which is actively transcribing or in a potentially active state in the cell. Actively transcribing genes are preferentially digested by DNaseI, a non-specific endonuclease. This DNase sensitivity of actively transcribing genes is dependent upon their association with the HMG14 and HMGl7 proteins. The association of HMG14 and HMG17 with active chromatin.was first shown by Weisbrod.and.Weintraub in 1979 (8). In erythrocytes, the globin gene is active and preferentially digested with DNase while the ovalbumin.gene is inactive and is not sensitive to this enzyme. In erythrocyte chromatin depleted with 0.35 M NaCl, the globin gene looses this preferential sensitivity while the ovalbumin gene remains unaffected. Reconstition of chromatin or mononucleosomes with the 0.35 M NaCl fraction or purified HMG14 and/or HMG17 restores the DNase sensitivity of the globin gene while the ovalbumin gene remains unaffected (7, 8) . The cellular source of the HMG14 or HMG17 used did not alter the specificity of the reconstitution. Full sensitivity to DNaseI is restored to active genes at about 1 mole HMG14 or HMG17 per 10-20 nucleosomes, a ratio equivalent to the amount of these HMG proteins found in the nucleus (7). In a similar manner, HMG14 and HMG17 were also shown to 34 specifically confer DNase sensitivity to nuclear RNA (nRNA) sequences as well as bulk active sequences (7,9). Nucleosome fractions containing active genes were found to be enriched for HMG14 and HMG17 and nucleosomes with bound HMG14 and HMG17 were found to be enriched in active, DNase sensitive sequences (12,32). HMG14 and HMG17 immobilized by crosslinking to agarose specifically retained actively transcribing nucleosomes when exposed to bulk digested chromatin (10) .- Quantitative analysis in chicken erythrocytes of the distribution of chromosomal proteins on transcribed chromatin shows a 1.5-2.5 higher density of HMG14 and HMG17 and a 2 fold lower density of histone H1 and H5 as compared to inactive sequences (36). Antibodies to HMG17 were also used to specifically isolate nucleosomes containing active genes.(11,33). In oligonucleosomes isolated with antibodies to HMG17, the distribution of HMG17 with respect to the coding region of specific genes was analyzed in detail. It was found that HMG17 is bound only downstream of the transcriptional start site (34). This correlates with.the observation that the coding regions of active genes are most sensitive to digestion by DNaseI, while regions upstream.are only moderately sensitive. All of these results suggest that HMG14 and HMG17 bind specifically to regions of chromatin containing active genes and indicate that active nucleosomes possess some feature that allows HMG14 and HMG17 to 35 specifically distinguish them.from the inactive nucleosomes. It is known that in active chromatin the core histones are hyperacetylated and there is a marked decrease in the amount of histone H1 bound. These changes and their consequences on the fine structure of nucleosomes may be involved in the specificity of binding of the HMG14 and HMG17 proteins. Indeed, HMG14 and HMG17 have been shown to partially inhibit histone deacetylase in active chromatin (51). Due to this well established association of HMG14 and HMG17 with active gene sequences, the influence of these proteins on the process of transcription was analyzed. Microinjection of antibodies to HMG17 into fibroblasts inhibited transcription as shown by significant reductions in nuclear incorporation of tritiated uridine (13). This suggests that the HMG proteins are required for transcriptional activity in the cell. The ability of these HMG proteins to function as direct transcriptional activators was studied in Saccharomyces cerevisiae cells (35). These cells express fusion proteins comprised of the lexA DNA binding domain fused to the HMG14 or HMG17 acidic C—terminus, which has features reminiscent of some transcriptional activators. This construct was unable to stimulate transcription from reporter constructs containing the lexA operator sequences upstream of the B-galactosidase gene. Thus, in this system at least, HMG14 and HMG17 36 do not function as direct transcriptional activators. modifications of the HMG14 and HMG17 proteins HMG14 and HMG17 are known to undergo several post -transcriptional modifications including methylation, acetylation, phosphorylation, poly(ADP) ribosylation and glycosylation. The functional significance,if any, of these modifications is unknown. These HMG proteins can be poly(ADP) ribosylated (38); in the chicken.proteins the putative sites for this modification are glutamine 81 in HMG14a and glutamine 70 in HMG17. While the biological significance of this modification are unknown, the ribosylation sites are located in the C—terminus of the molecules and may be involved in interactions of the HMG proteins with the nucleosome and/or the core histones. Studies with trout H6 protein show there is a preferential localization of ribosylated HMG proteins in DNase sensitive chromatin (37). HMG14 and HMG17 can be glycosylated, and in-vivo labeling experiments show these proteins can incorporate fucose, galactose, mannose and NLacetylglucosamine. Most of these oligosaccharide linkages are of the Neglycosidic type (38). The function of these modifications are unknown, yet it has been shown that gylcosylated HMG proteins bind preferentially to the nuclear protein matrix in mammalian cells (39). The nuclear matrix is the site of both RNA transcription and DNA 37 replication which is consistent with a role of these proteins in actively transcribing chromatin. Phosphorylation of HMG14 in mammilian cells can occur at two sites, serine 6 and serine 24. These residues are located in two highly conserved regions of the protein. Phosphorylation at these sites was shown to be hormonally regulated by thyroid stimulating hormone in-vivo and can be mumicked by forskolin and cAMP analogues in-vitro (40). HMG17 is also phosphorylated but does not show this regulation. Phosphorylation of HMG14 reduces its ability to interact with nucleosomes by diminishing interactions with DNA. This could be an important regulatory mechanism for HMG14 binding to nucleosomes and illustrates an important biological difference between HMG14 and HMG17. ' Tissue specificity and developmental regulation of HMG14 and HMGI7 Antibody injection experiments have localized HMG14 and HMGl7 to the nucleus in human fibroblast cells (41). There appears to be no functional difference in the HMG proteins produced in different tissues as HMG14 or HMG17 from chicken brain cells can reconstitute DNase sensitivity on erythrocyte chromatin (7). The cell cycle regulation of HMG17 mRNA has been analyzed in.HeLa cells (43). The level of HMG17 mRNA in these cells is high and does not correspond to the amount of protein.present 38 in the cell. The level of HMG17 mRNA is much higher than the level of the message for actin which is a much more abundant protein in the cell. This suggests that the HMG17 is rapidly turned over, however this conclusion is disputed by other evidence. HMG17 mRNA is present throughout the cell cycle with a sharp rise at the beginning of S-phase, followed by another rise near the end of S-phase. This rise in the rate of transcription of the gene is not coupled to DNA synthesis as inhibition of DNA sythesis does not affect HMG17 mRNA synthesis (43). Interestingly, the level of HMG17 protein present in HeLa cells does not change noticeably with fluctuations in the mRNA levels (15). Isolation.of HMG14 and.HMG17 proteins from various organs of the chicken show that in organs with a higher proportion of replicating cells, the amount of HMG17 is much higher than HMG14 while in transcriptionally active organs with a very small proportion of replicating cells, the levels of HMG14 and HMG17 are low and roughly equal (42). The relative levels of HMG14 and HMG17 proteins varies in the different cells tested. In chicken erythroctyes and transformed quail fibroblasts, HMG14a is more abundant than HMG17, while in many human cell lines, HMG17 is more abundant than.HMG14 with slight variations in the ratio depending upon growth conditions (50). The expression of these two proteins does not seem to be coordinately regulated, as changes in the 39 level of one does not effect the level of the other. In a chicken lymphoid cell line, that has the single copy gene for HMG17 inactivated at both allelles, there is no HMG17 protein present in the cell yet the level of HMG14a does not seem to be affected (Yi Li, personal communication) . However, recent results from our lab suggest that HMG14 protein levels may be down regulated in transformed quail cell lines ectopically expressing high levels of chicken HMG17 (unpublished results). There is evidence that HMGl4 and HMG17 mRNA synthesis may be coupled to the differentiation state of the cell. In myogenesis, myoblasts differentiate into myotubes. During mouse myogenesis, the HMG14 and HMG17 mRNA levels and protein synthesis rate decrease to 20% of that seen in myoblasts, while the levels of the protein in the cell remains unaltered. These fluctuations were not coupled to DNA synthesis (44). Similar results were shown in rat cells, with HMG14 and HMG17 mRNA and protein synthesis rates in proliferating myoblasts significantly higher than in nondividing myotubes (45). To investigate the importance of the downregulation of the HMGs in myogenesis, exogenous human HMG14 expressed in mouse myoblasts was found to block the differentiation of these cells to myotubes (48). This suggests that proper regulated expression of the HMG proteins may be required for some aspects of cellular differentiation. 40 This is consistent with a role of the HMG14 and HMG17 proteins in gene regulation. During erythropoiesis in chicken, the levels of the HMG proteins and their mRNAs also change as the cells develop from embryonic to adult stages (46,47). Embryonic erythroid cells have significantly higher levels of HMG14 and HMG17 mRNAs than adult cells. In addition, HMG14a, the major chicken HMG14 protein, is expressed primarily in embryonic and developing cells while HMG14b is preferentially expressed in definitive cells. The developmental regulation of the HMG14 and HMG17 proteins was also analyzed in osteoblast cells that can be induced to differentiate into osteocytes and in promyelotic leukemia cells that can be (induced to a post-proliferative state (49) . The results were consistent with the other reports in that the HMG proteins were downregulated as the cells progressed into their differentiated states. It is not clear whether the switch in HMG expression levels is a prerequisite or a consequence of differentiation. HMG14 and HMGl7 cDNAs cDNAs for HMG14 and HMG17 have been isolated and sequenced from the chicken, human, and mouse. Two types of HMG14 have been isolated from the chicken, HMG14a, which is the major chicken HMG14, and HMG14b, a slightly smaller protein, more closely related to the human HMG14. The 41 overall structure of the cDNAs for the HMG14 and HMG17 proteins are similar with clear sequence distinctions between the subfamilies. All the cDNAs for HMG14 and HMG17 contain a short 5' untranslated sequence with no sequence similarity among the clones. The coding region of these cDNAs have stretches of high similarity. Among the HMG14 cDNAs, the sequence similarity in this region ranges from 61% to 77%, while the HMG17 cDNAs show a higher degree of evolutionary conservation with sequence similarites ranging from 85% to 93%. The 3'-untranslated region of all the HMG14 and HMG17 cDNAs is unusually long, A+T rich, and is highly conserved in certain sections among and between the subfamilies. In the HMGl4 subfamily, the first 250 nucleotides adjacent to the coding region shows similarities ranging from 48% to 86%, with the more distal segment being less conserved. In the HMG17 subfamily, the first 150 nucleotides adjacent to the coding region show sequence similarities ranging from 85% to 95%, with the similarity being less pronounced in the distal portion of the clones. Between the HMG14 and HMG17 subfamilies, this portion of the 3'-untranslated region is also highly conserved. The functional significance of the high level of conservation seen in the 3'-untranslated portion of these cDNAs is unknown, but this region may be important for gene regulation, message 42 stability, or processing. A dot matrix analysis comparison of the human HMG14 and HMG17 and the chicken HMG14a and HMG17 cDNAs is shown in figure 6 (53). Chicken HMG14a cDNA The cDNA for chicken HMG14a was isolated by screening a chicken liver cDNA library with synthetic Oligonucleotide pools whose sequence was derived from the partial amino acid sequence of chicken HMG14 (53). The sequence of the chicken HMG14a cDNA is 900 bp in length. It contains 104 bp of 5'-untranslated sequence and a long 3'—untranslated sequence of 445 bp. The coding region codes for a protein of 104 amino acids. The complete cDNA sequence for this clone is shown in figure 7. Chicken HMGI? cDNA The cDNA for chicken HMG17 was isolated by sceening a chicken liver cDNA library with synthetic Oligonucleotide pools derived from the complete amino acid sequence of chicken HMG17 (53). The cDNA is 1360 bp in length. It contains 177 bp of 5'-untranslated sequence and a 843 bp 3'-untranslated region. The coding region predicts a protein of 89 amino acids. The complete cDNA sequence for chicken HMG17 is shown in figure 8. 43 Figure 6 - Dot matrix comparison of HMGI4 and HHGI7 cDNA sequences Dot matrix comparison of HMG14 and HMG17 cDNA sequences (53). (A) Comparison of human HMG17 cDNA (X-axis) to chicken HMG17 cDNA (Y-axis). (B) Comparison of human HMG14 cDNA (X axis) to chicken HMG14a cDNA (Y-axis). (C) Comparison of chicken HMG14a cDNA (X-axis) to chicken HMG17 cDNA (Y-axis). A window of 10 nt residues was used with a 50% identity required for a positive result. 44 - 0’0..." e 5. 0.0... e o ..... .34.»... u .r ........ an... . . ”a“. .n on.“ 1000 o e. u .0 e 00 e. .- utd- - o 0. no. on o o O . e C's-o ' a... e . a. u . I .0. o a o no . u 0 Q...- eouo 0 same \ o o I I “no .00. eq-uvunw ’0. of ”0 coo 10m 1200 400 100200300 700 400500600 Figure 6 45 Figure 7 — Nucleotide sequence of chicken HMG14a cDNA Nucleotide sequence of chicken HMG14a cDNA. The complete sequence of the chicken. HMG14a cDNA is shown with the translated amino acid sequence beneath. Seven bases of EcoRI linkers are at each end of the clone. 46 GAAITCCGTC CCCTICCTCA GGACGCTCGA WC WC‘CCTICCIAIT MCI WCICIAITIGC AGTCAACTAT W crarccccaa WWWWWMWWW 2 sArgLysAla ProAlacluG lyGluAlaLy sGluCIuPro LysArgArgS erAlaArgLe AICTGCIAAA CCIGCTCCGC CIAMCCGGA GCCAAAGCCC MMAGGCAG CACCTAAGAA 22 uSerAlaLys ProAlaProP roLysProGl uProLysPro LysLysAlaA laProLysLy - W W ' W641 GCAMCGATA W W 42 sGluLysAla AlannAspL ysLysGluAs pLysLysAla AlaThrLysG lyLysLyscl WC MAGACGAAA W GGATGCAMA CW6 ACTCTGMM 62 yAlaLysGly LysAspGluI hrLysclncl uAspAlaLys GlucluAan isSercluAs WC MMCTMTG AGGCACCAGC TGCTGAAGCA TCTGAIGATA mom 82 nclyAspThr Lys‘nxrlunc luAlaProAl aAlaGluAla SerAspAspI. yscluAlaLy GICCGAGTAA TGTTAACCCT GCCC'IA‘IATC ICCA‘ICA‘ITI GGTAICCGTA CCTCCATGCI 102 sSerclufi'k GTATTGTTAA WA TATITTIATC mm MTGCAGG rrn'rmcc mm TTATGGMCA TCTTCATCTC 6mm GAATIAMTC CCTAACAMC W MMCAMM MMTCATTG TTTIAMTIT GIGATTGTM TAGTTTGTAT GGTACA‘IGGA MGAATMGT GGTGGTAGCT mGACTICT GTCAGTGTGT CCCTT'ITIGT GTMGTCATG CTTACAGACT TCAGATTTTA ATTTTACCCT TGTATGTGTT GTATGGTT'IC m GAGGTCTCAA MCAGAIAAC TGTGTTAMC ATTCCAG‘IGG TTCTGTGGGT W WA GCTATTT'ICA TGMAAMM AW MCCGAATIC Figure 7 120 no 21.0 300 360 4.20 no 540 660 i '720 780 900 47 Figure 8 - Nucleotide sequence of chicken HHGI7 cDNA Nucleotide sequence of chicken HMG17 cDNA. The complete sequence of the chicken HMG17 cDNA is shown with the translated amino acid sequence beneath. Seven bases of EcoRI linkers are at each end of the clone. 48 GAAITCCGCA GCCAGCGCAG GGAGCCGGCC CCTGGGCCCT cccccccrrc TCGCCGCCAC cccccrcccc TCGCTCTCTC CCICCICGCA cm W AGATACCAAG , l ProLysArgL yaAlaGluGl yAspThrLys W 006W ATCTGCTAM 21 cmwgs erAlaArgLe uSerAlaLys W CTOCAAAGAA W ‘1 LysLysAlaA wroLysLy sSerGluLys GCTGGCAAGG W COW 61 AlaGlyLysG luclyAsnAs nProAlaGlu AAAGCCGAAG GTCCTGGTGA IGCCAAGTAA 81 LysAlaGluG lyAlaGlyAs pAlaLysm CTGGTGACTG W MTACTATIT ICTITIACTT ITTTIAAGCT AICTTCTTAG GGGGGGGGCA GTGGGACAAA CCTCACTIAA ITACCCCTTC CCAGTTTTTT AGAAGGACTC GTGCIGCACA CCTCTTCCCT ITTGTGGACC CCTGTTGCCA AC'ITCAGMC TGCAGTTTGC CCTTTTIGCC IAGAGCCTAI CACTCCGAAA ACICTAAAIC CAITGTCAGG TGATCTGGAC CIAAAAGGAG CTCCATTTCC TCTTTCAIAI TTTIANTTTT TCCTCGCAAA GCIAGGGTAG AAICTCAAAC,ICTCCGCCCI CACTCIAAAC AITIGICGGT ITIKIAGCAA CCTTIATGTT CAGTTCTTCI AAAAIGTTGC AGAITCIAGC IAIGAAAAGT ACCTTIAAIA AAGCTCGAIA AAAAAAAAAA AAAAAAAAAA‘AAAAAAAAAA Figure 8 GCCAGCCCCG CCGCGCCGCC CCGCTC'ICCC WOO 0066100666 6666666666 CAACACAOGC ACGCGCCGCC CGGAGCTATG GGCGATMGG CCAMGTIAA WC“ GlyAspLysA laLysVaILy sAspGluPro CCTGCCCCTC OGAAGCCAGA meet 60 120 180 250 ProAlaProP roLysProGl uProLysPro ’ , W W W ValrroLysG lyLysLyscl yLysAlaAsp AATGGAGATG CW CCAGGCACAG AsnclyAspA laLysThrAs pGlnAIacln MTGTGTGAA TI'ITTGATAA CTGTGTACTT TTIATCAAGT ITINIAACAA.TCCAGAAIII CACACAGACC GCITTCTTCT TCTGTTTTGA TCTGTTTCTI GGAACCTAAA.TTTIAAAAGT TTCCIAAATC GAGCAGGAAB GGKTTCCTIC GCATCAGACT GAACGGAAGC ICCCGAGAIG AGTCCCCTCT GCGTTTCCTT ICAICCCCIC IACAGCAGAC AIGGCATGTT GGGACTCACC TTCTCGTCTC IAAITTCGGA IAIAAIAGCT TGTAGAICIA CAGATTAAGG AAICTCCAGT ATTTCTGAAG AGTICTTAAA CAACAICCIA AITTCCCTCI ACAAGTAIAC AAAAAIGAAG TCGGTAGTCC ATGAAGGGAG GGGAGTTTGA CCATCTCCTG CCIAAAITAC CATGAIICTT CGGTTTCCCI TCGAAAAAAA.AAAAAAAAAA AACGGAAITC _ - 360 (+20 480 660 720 780 900 960 1020 1080 1140 1200 1260 1320 1360 49 HMG14 and HMGI7 chromosomal genes The chromosomal genes for HMG14 and HMG17 have been cloned and sequenced in the human and chicken (53-59). Southern analysis shows that the human genome contains 35 -50 HMG17 gene copies and 60-90 HMG14 gene copies (60,61). Most of these are believed to be retropseudogenes making HMG14 and HMG17 the largest known human retropseudogene families. The structure of the human genes are similar to the chicken genes which are described below. Comparison of the genomic structure of the HMG14 and HMG17 genes suggests these genes evolved from a common ancestor. The functional human gene for HMG17 has been mapped to band 1p36.1 and the many pseudogenes are disbursed over several chromosomes (55). Interestingly, the functional human gene for HMG14 maps to chromosome region 21q22.3 which is the region associated with the pathogenesis of Down syndrome (63,64). Down syndrome is characterized by extra copies of this region on the chromosome. RNA and protein analysis from mouse trisomy 16 embyros, a mouse model for Down syndrome, shows elevated levels of HMG14 present. It is not known if elevated levels of this chromosomal protein contribute to the etiology of the disease. Chicken HMGl4a chromosomal gene The chicken genome contains one gene each for HMG14a and 50 HMG14b. HMG14a is the major HMG14 expressed in avian cells with HMG14b, the mammalian homologue, being expressed at low levels. It is not known if there is a homologue for HMG14a in mammalian cells. The chromosomal gene for HMG14a was isolated by screening a chicken library with the HMG14a cDNA (56). Positive clones were picked and analyzed by restriction enzyme and sequence analysis. The HMG14a gene spans approximately 10 kb and is single copy in the chicken genome. The gene contains 7 exons and 6 introns, the exons being labeled exon 0 through exon-6. Exon 0 contains about 126 bp of S'non-coding sequence. Exon 1 contains the rest of the 5' untranslated region, the ATG translational initiation codon and the first 4 codons of the protein sequence. Exons 2,3,and 4 are small at 30, 30 and 51 bp, respectively, and exon 5 is somewhat larger at 144 bp. These exons contain all coding region sequence. Exon 6 is the longest exon and contains the last 15 codons of the coding sequence, the termination codon, TAA, and the entire 3'-untranslated region of the gene. The exon-intron boundries have been mapped for this gene and the splice acceptor and splice donor sites are similar to those found in other nuclear genes (56). The HMG14a promoter region has been sequenced and was found not to contain the consensus TATAA or CCAAT elements 51 found in.many eukaryotic gene promoters; however a number of putative SP1 binding sites were located therein. The promoter region is very G+C rich at 76% and contains 7 HpaII sites (CCGG) (56). These features are characteristic of housekeeping genes which are transcribed constituitively in most cells. The 3'-untranslated region of the HMG14a gene contains 4 copies of the canonical polyadenylation signal AATAAA. Primer extension analysis and SI mapping verify that the HMG14a gene is transcribed into multiple mRNAs arising from two or more transcriptional start sites with alternative splicing and utilization of two or more polyadenylation signals (59). The chicken HMG14b gene is similar in structure to that of HMG14a. The promoter region contains no consensus elements other than the SP1 binding sites and is very G+C rich. Also, like the HMG14a gene, it contains multiple polyadenylation sites (56). Aside from very different protein coding sequences, the HMG14b gene differs from HMG14a in that exons 2 and 3 are fused. These exons code for the most highly conserved portion of the HMG14 and HMG17 protein. Chicken HMGI? chromosomal gene The chicken genome contains one single copy gene for HMG17. The chromosomal gene for HMG17 was isolated by 52 screening a chicken genomic library with the chicken cDNA for HMG17 (56). Positive clones were analyzed by restriction enzyme analysis and DNA sequencing. The HMG17 gene spans less than 4 kb in the genome, much less than the gene for HMG14a. The gene contains 6 exons and 5 introns. Exon 1 contains all of the 5' untranslated sequences, the ATG translational initiation codon and the first 4 amino acid codons for the protein. Exons 2,3,4 and 5 are small, at 45 bp, 30 bp, 51 bp and 96 bp, respectively, and contain protein coding sequences. Exon 6 is much larger than the other exons and contains the last 33 codons of protein sequence, the TAA termination codon and all of the long 3' untranslated region of the gene. The extensive similarities in genome structure of the HMG14a and HMG17 genes suggest that they evolved from a common ancestor. The exon-intron boundries have been mapped for this gene and the splice donor and splice acceptor sites are similar to those found in other nuclear genes. The promoter region for the HMG17 gene has been sequenced and was found to contain many common promoter elements, unlike the HMG14a gene. The promoter region of HMG17 contains two TATAA boxes 31 and 41 bp upstream of the initiation site. A CCAAT element is found 61 bp upstream and the G+C content of the region is high at 75% with 4 putative SP1 binding sites. The existence and location of 53 these promoter elements is typical of many eukaryotic genes. The 3' untranslated region of the HMG17 gene contains one copy of the canonical polyadenylation signal AATAAA 27 bp upstream of the polyadenylation site. Chapter II Materials and Methods 54 55 Chapter II materials and Methods 1. Subcloning HMG14a and EKGI7 wild type and.mutant cDNAs into expression vectors HMG14a and HMG17 cDNAs were originally cloned into the EcoRI site of plasmid pCla12 in both the sense and antisense orientations. pCla12 is a 2 kb adapter plasmid designed to aid in cloning fragments into expression vector TFANEO (70). The polylinker of pCla12 is flanked by ClaI sites, allowing the cDNAs to be excised with ClaI and ligated into the CiaI site of expression vector TFANEO (71). TFANEO is a 7.2 kb expression vector used in our experiments. The expression cassette of TFANEO consists of two LTRs derived from the Schmidt-Rupin avian Rous sarcoma virus that provide transcriptional and polyadenylation signals. A unique ClaI restriction site lies between the two LTRs. This vector also contains the ampicillin resistance gene and the neomycin gene driven by a chicken B-actin promoter for selection by G418 in tissue culture. All ligations were carried out at 15° overnight using T4 DNA ligase and buffer from.New England Biolabs (NEB). When neccessary, blunt ends were created by filling in the ends of the fragment and vector with the Klenow fragment of DNA polymerase I (NEB) using standard procedures (72). Free 56 ends of fragments and vectors were dephoshporylated with calf alkaline phosphatase from Boehringer Mannheim (BM) using standard procedures (72). Ligation reactions were transformed to E.Cb1i DHB by standard procedures and plated onto Luria -Bertani medium (LB) plates containing ampicillin for plasmid selection (72). Colonies were picked and plasmid DNA was isolated by alkaline lysis and tested by restriction enzyme digests for presence of insert and orientation. Most of our restriction enzymes were purchased from NEB or BM. The orientation of the HMG cDNAs in TFANEO was determined by restriction enzyme analysis as follows. Within the pCla12 polylinker, there is a BamHI site located 3' to the 5' ClaI site and 5' to the EcoRI site, such that the BamHI site is always 5' to the cDNA insert when the insert is cloned in the sense orientation in pCla12. Additionally, there are two BamHI sites in TFANEO located 5' to the ClaI cloning site in this vector. These sites are located 600 and 1000 bp upstream of the ClaI site. When digesting TFANEO containing an HMG cDNA insert, a 400 bp BamHI band is always present and a 600 bp band will be present when the insert is in the sense orientation. If the insert is in the antisense orientation, there will be a 400 bp BamHI band along with a band of 600 bp plus the size of the cDNA insert. This cloning strategy is illustrated in figure 1. 57 Figure 1 - TFANEO subcloning strategy The subcloning strategy to move cDNA inserts from.plasmid pCla12 to plasmid.TFANEO is shown. This is described in detail in the text. 58 omz: £83 32 2.3 0.5 Etobmcok .358 259.2. Becca D D D D 3 .J ‘1 A: .. In <20 9:52:00 :85 popcotfi 205m 220m. 3 D Q: ommsu scam . 25 Iowa. 320 OS .Sp 2.... 58.8..» Figure 4 91 Figure 5 - Oligonucleotides used for sequencing mutants in the HMG14a and HMGI7 cDNAs The Oligonucleotide primers used for sequence verification of site-specific mutations in the HMG14a and HMG17 cDNAs are shown and their locations are indicated. Oligo name 14seqpr1(AS) 14seqpr2(AS) 14seqpr3(AS) A' (S) D' (AS) 92 Sequence 5' - 3' TGGCTCCTTTCTTCCCTT GCTGGTGCCTCATTAGTTTTGG GCATCCTCTTTGTTTAG AGGCGTATCACGAGGC TCTGACTTGAGCGTCG Figure 5 Location of 3' end nt 287 in cDNA nt 368 in cDNA nt 320 in cDNA nt 2791 in pCla12 60 bp 5' of polylinker nt 1035 in pCla12 135 bp 3' of polylinker 93 PCR generated fragments as template DNA and the two outside primers, A' and D'. This produces a DNA fragment that includes the entire cDNA with an internal mutation as well as sequences outside of the convenient cloning sites, EcoRI and ClaI. This fragment is then digested with EcoRI, releasing the mutated cDNA and two small fragments from either end of sizes 60 and 130 bp, which can be visualized on an agarose gel to verify digestion. This procedure eliminates the:need.totdigest the PCR.product close to the ends of the fragment which can produce uncohesive ends. An illustration of this PCR.based mutagenesis strategy is shown in figure 6. The mutated cDNA was then subcloned back into pCla12 for sequencing verification of the mutation. The mutated cDNA was then subcloned into the appropriate expression vector for use in tissue culture experiments. 2. Production of deletionumutations in.HMG14a and.HMG17 cDNAs Among the HMG14 and HMG17 cDNAs studied from various species, the 3' untranslated region contains a region which is the most highly conserved portion of the cDNA. A dot matrix analysis of the chicken and human HMG14 and HMG17 cDNAs located this region in the chicken HMG14a and HMG17 at nt 500 - 550 (59). Although the significance of this conservation is unknown, these sequences may be important for 94 Figure 6 - PCR mutagenesis reaction scheme A schematic of the procedure used for making site-specific mutations in the HMG14a and HMG17 cDNAs is shown. The individual steps involved are described in detail in the text. 95 ~moow axoow toaom> co_mmotoxo m§§§§§§§§§§§§§§§§§§§§§§§ oa mac—093m 9 women ocom ooESE < website 23230 to oEmE 9:82 9:38 cozotammt 2:; B36 .moow amoow _ _ J 12\ N t .O _ u .< .D..< .3058 02.330 oz; 2:3 mom macoemmt mom 02;. mam—om“ .Q Eoow Eoow II. _ Ik? _ it <|l 1 .< Amt m A Im .aoom \ /N\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\A\ ||<|u .IIII .< toaoo> 9.30.0 .motmts to 969 N Figure 6 96 regulation of transcription, proper mRNA polyadenylation and/or mRNA structure and stability. To investigate these possibilities, this region was deleted in both the HMG14a and HMG17 cDNAs. In HMG14a, this conserved 3 'untranslated region was deleted by digesting pC14cw with the restriction enzyme RsaI, which cuts at postions 470 and 664 in the HMG14a cDNA” The internal 196 bp fragment was removed and the plasmid was ligated back together. The result is removal of the highly conserved region and some surrounding sequences, leaving 276 bp of 3' untranslated sequence. In HMG17, this region was deleted by digesting pC17cw with enzymes RsaI and DraI, which cut at positions 477 and 655, respectively, in the cDNA, removing the 178 bp fragment and religating the plasmid back together. This leaves 725 bp of 3' untranslated sequence in the cDNA. These two deletion mutants were then subcloned into expression vectors for use in tissue culture experiments. The location of the 3"untranslated region.deletions are shown in figure 7. 3. Production of mutations in HMGl4a and HMG17 altering the site of poly(ADP) ribosylation The HMG14 and HMG17 proteins undergo a number of posttrancriptional modifications. Among these is poly(ADP) ribosylation. The effects of this modification on the 97 Figure 7A - DNA sequence and location of mutations in the HMG14a cDNA The complete cDNA sequence of chicken HMG14a. The 3' untranslated region deletion defined by RsaI at nt477 and nt 664 is underlined. The location of the nt mutation designed to alter the poly(ADP) riboslyation site is starredl The mutation is G-A and changes the amino acid from Glu(GAA) - Arg(AGA). 98 HMGl4a cDNA GAATTCCGTCCCCTTCCTCAGGACGCTCGAAAACAGTTTCTCGGCGGTTCCCTTCCTATT 1 TTTTACACCTCTCCCGATCTCTCTATTTGCAGTCAACTATTAAGGTGCAACTATGCCCAA 61 AAGAAAGGCTCCAGCTGAAGGCGAGGCGAAGGAGGAGCCAAAGAGAAGGTCGGCCAGACT 121 ATCTGCTAAACCTGCTCCGCCTAAACCGGAGCCAAAGCCCAAAAAGGCAGCACCTAAGAA 181 AGAAAAGGCAGCAAACGATAAAAAGGAAGACAAAAAGGCAGCAACAAAAGGGAAGAAAGG 241 AGCCAAAGGCAAAGACGAAACTAAACAAGAGGATGCAAAAGAAGAAAACCACTCTGAAAA 301 + TGGAGATACCAAAACTAATGAGGCACCAGCTGCTGAAGCATCTGATGATAAGGAAGCCAA 361 GTCCGAGTAATGTTAACCCTGCCCTATATCTCCATCATTTGGTATCCGTAQQIQQAIQQI 421 TA TTAACAGA AGGAATATTTTTAT “ ‘ATTTTATAAA . ‘_L A 481 A AATTT“TTA .LAA ‘T TTCAT .L AA ‘ .LAATT“‘TC T“_“‘ 541 Alta-‘fiiACAAAAAAAAAAAAATAIL Al_;_ '1 ALTAATA TAT 601 QQTACATGGAAAGAATAAGTGGTGGTAGCTTTTGACTTCTGTCAGTGTGTCCCTTTTTGT 661 GTAAGTCATGCTTACAGACTTCAGATTTTAATTTTACCCTTGTATGTGTTGTATGGTTTC 721 TTAAAGTGGGGAGGTCTCAAAACAGATAACTGTGTTAAACATTCCAGTGGTTCTGTGGGT 781 TGCTTTTATAAAGAAGGTGAGCTATTTTCATGAAAAAAAAAAAAAAAAAAAACGGAATTC* 841 900 Figure 7A 99 Figure 7B - DNA sequence and location of mutations in the HMG17 cDNA The 3' untranslated region deletion defined by RsaI at nt 477 and DraI at nt 655 is underlined. The location of the nt mutation designed to alter the poly(ADP) riboslyation site is starred. The mutation is A-C and changes the amino acid from Glu(GAA) to Ala(GCA). 100 HMGI? cDNA GAATTCCGCAGCCAGCGCAGCGAGCCGGCCGCCAGCCCCGCCGCGCCGCCCCGCTCTCCC 1 CCTCGGCCCTCCCCCGCTTCTCGCCGCCACCGAGCGAGCCCGGCTGCCCGCCCCCGCCCG 61 CCCCCTCCGCTCGCTCTCTCCCTCCTCGCACAACACACGCACGCGCCGCCCGGAGCTATG 121 CCGAAGAGAAAGGCTGAAGGAGATACCAAGGGCGATAAGGCCAAAGTTAAGGATGAGCCA 181 CAACGGAGATCGGCAAGGTTATCTGCTAAACCTGCCCCTCCGAAGCCAGAGCCTAAACCT 241 AAAAAGGCAGCTCCAAAGAAGAGTGAGAAGGTGCCCAAGGGAAAGAAGGGGAAAGCTGAT 301 , + GCTGGCAAGGAGGGAAACAACCCTGCAGAAAATGGAGATGCCAAAACAGACCAGGCACAG 361 AAAGCCGAAGGTGCTGGTGATGCCAAGTAAAATGTGTGAATTTTTGATAACTGTGTAQTT 421 'LJI A TACA. AAATA ATTTTTTA " TTTTA AACAA _ ' AATTT 481 L rill} AA .ATG TTA ‘CA 'GA G' TTG A 541 QQQQQQQQQAGTQQGACAAACGTCACTTAATQTGTTTQIIQQAACCTAAATTTTAAAAGT 601 TTACCCCTTCCCAGTTTTTTAGAAGGACTCTTCCTAAATGGAGCAGGAAGGGATTCCTTC 661 GTGCTGCACACCTCTTCCGTTTTGTGGACCGCATCAGAGTGAACGGAAGCTCCCGAGATG 721 CCTGTTGCCAACTTCAGAACTGCAGTTTGCAGTGCCCTCTGCGTTTCCTTTCATGCCCTC 781 Figure 7B 101 CCTTTTTGCCTAGAGCCTATCACTCCGAAATACAGCAGACATGGCATGTTGGGACTCACC 841 ACTCTAAATGCATTGTCAGGTGATCTGGACTTCTGGTGTCTAATTTGGGATATAATAGCT 901 CTAAAAGGAGCTGCATTTCCTCTTTCATATTGTAGATCTACAGATTAAGGAATCTGCAGT 961 TTTTAATTTTTCCTCGCAAAGCTAGGGTAGATTTGTGAAGAGTTGTTAAACAACATGCTA 1021 AATGTGAAAGTGTCCGCCCTCACTCTAAACATTTCCCTCTACAAGTATACAAAAATGAAG 1081 ATTTGTCGGTTTTATAGCAACCTTTATGTTTGGGTAGTCCATGAAGGGAGGGGAGTTTGA 1141 CAGTTGTTGTAAAATGTTGCAGATTGTAGCCCATGTCCTGCCTAAATTACCATGATTGTT 1201 TATGAAAAGTACCTTTAATAAAGCTGGATACGGTTTGGCTTGGAAAAAAAAAAAAAAAAA 1261 AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAACGGAATTC 1321 1360 Figure 7B (con.) 102 function of these HMG proteins is unknown but the putative site of this modification has been localized to glutamic acid residues 81 in HMG14a and 7.0 in HMG17. Poly(ADP) ribosylation has been implicated in crosslinking of other chromosomal proteins such as the histones and modification of these residues may lead to alterations in the proteins ability to functionally interact with the nucleosomes. These residues have been altered by Oligonucleotide directed in-vitro mutagenesis by others in our lab. In HMG14a, this glutamic acid residue (GAA) was changed to arginine (AGA), and in HMG17 to alanine (GCA). The locations of these mutations are shown in figure 5. DISCUSSION For the production of site-specific mutations in the HMG14a/17 cDNAs, the in-vitro mutagenesis procedure proved to be the most expedient and efficient. Once a stock of single -stranded uracil containing DNA was isolated for each clone in M13, the procedure can be completed in a matter of hours using standard reagents and techniques. The frequency of obtaining the desired mutation was 50% or higher. For each mutagenesis reaction performed, 4 colonies were picked and their DNA was sequenced over a region spanning approximately 200-300 bp within the cDNA and including the mutated region. Consistantly, at least 2 of these clones contained the 103 desired mutation. Undesired mutations, presumably produced from mistakes made during replication.by the DNA.polymerase, were noticed at a very low but detectable level. Out of approximately 34 clones sequenced, 3 mismatches were detected in the region sequenced, 2 in one clone and 1 in another. Several mutations were produced using the PCR based strategy. This procedure is slightly more time consuming as it requires two consectutive PCR reactions and the reaction conditions must be optimized for each set of mutagenic primers used. In addition, the amplified bands frequently required gel isolation/purification prior to the second amplification and prior to subcloning with the final DNA fragment. Subcloning with PCR generated fragments is often problematic, in part due to the periodic addition of extra nucleotides on the fragment ends. This problem can be avoided by restriction enzyme digestion of the final product (provided there are convenient sites available) designed to cut off the ends and produce a fragment with the appropriate ends for subcloning. This was accomplished.with.our clones by using the EcoRI sites located internal to the external primer binding sites. The location.of the external primers intpCla12 were chosen such that digestion with EcoRI releases two end fragmnets of detectable size on a mini gel for verification of digestion. The restriction enzyme digest was often done directly in the final PCR reaction mix. For efficient 104 subcloning of the resultant fragment, it is neccessary to thoroughly extract the final reaction mix with phenol/chloroform followed by an ether extraction to completly rid the DNA fragment of excess nucleotides and primers, Taq polymerase (which may add nucleotides to fragments ends) and all traces of phenol (which can prohibit subsequent ligation). The efficiency of~this PCR based mutagenesis strategy was 100% as detected by sequencing or restriction enzyme digest of the final subcloned product. Although only a few clones were sequenced, no undesired mutations were detected in the region sequenced. Chapter IV Expression of chicken HMG14a and HMG17 proteins in tissue culture cells 105 106 Chapter IV INTRODUCTION Chicken chromosomal binding proteins HMG14a and HMG17 bind to and interact with nucleosomes in regions of active or potentially active chromatin. While much research has focused on the specific physical interactions involved between the HMG proteins and the nucleosome, very little is known of the cellular function of this binding. We have chosen to address this question.by ectopically expressinngild type and mutant chicken HMG14a and HMG17 proteins in abnormal amounts in tissue culture cells and determining the resultant effects on the cellullar phenotype and the structure of the cellular chromatin. Each nucleosome contains two potential binding sites for the HMG14/17 proteins yet the amount of these proteins present in the cell is regulated such that only about one out of every ten nucleosomes actually has HMG14/17 bound. Overexpression.of wild type and.mutant HMG14/17 proteins may cause the exogenously expressed proteins to occupy new sites in cellular chromatin and, in the case of mutant proteins, to outcompete the endogenous HMG for binding sites on the nucleosome. Any effects of this aberrant binding or other influences the exogenously expressed protein may have on the cell phenotype and/or its chromatin structure can.then be observed. 107 As with many cellular proteins, the level of HMG14/17 present in the cell may be regulated, Bustin.et al. attempted to transiently express exogenous human.HMG14 and HMG17 in.COS cells and reported that while relatively high levels of mRNA could.be detected, HMG14/17 protein levels were limited to 2 -4 times that seen.in untransfected.cells (65). It is unclear if this level of exogenous HMG14/17 expression is a result of cellular regulation or, rather, of the limitations of the system used. We have undertaken a series of experiments designed to over express chicken HMG14a and HMG17 proteins for the first time in transformed quail fibroblast cells. 108 RESULTS Transfection of QT6 tissue culture cells QT6 cells are a chemically transformed quail fibroblast cell line used in our expression experiments (67). This cell line was chosen for these experiments due to our familiarity with its properties, and more important, its ease of transfection. There are very few immortalized, easily transfectable avian cell lines available. Transfections of QT6 cells were performed using standard calcium phosphate transfection procedures. The transfection efficiency using these cells is consistently high. Routinely, 10 ug of DNA per 10 cm tissue culture plate with 106 cells gave 500-1000 stable colonies following selection. In some instances, it was neccessary to reduce the amount of DNA transfected to ensure isolation of single colonies. Single colonies were isolated by picking well isolated colonies from the transfected plates. Transfection using expression vector TFANEO TFANEO is an expression vector designed by Federspiel et a1 (71). The wild type and mutant cDNAs for chicken HMG14a and HMG17 were cloned into a ClaI restriction site located between two long terminal repeat (LTR) sequences derived from the Schimdt-Rupin avian Rous sarcoma retrovirus . The upstream LTR provides promoter and enhancer functions while the 109 downstream.LTijolyadenylates the resultant transcript. This vector also contains a neomycin gene driven by the chicken B -actin promoter for selection by G418 in culture. A map of TFANEO is shown in figure 1. The wild type cDNAs for HMG14a and HMG17, deletion mutants 14-R and 17-R, and site-specific:mutants 14A1, 14A4, 14C1 and 17B6 were cloned into TFANEO, as described in chapter II and III, and transfected into QT6 cells. Stable pooled and single colonies were isolated from these transfections. Northern and western analysis were used to determine the level of HMG14a or HMG17 mRNA or protein ectopically expressed in these clones. Northern analysis of ENG14/17 transfected cell lines Figure 2 shows a northern analysis of total mRNA isolated from stable colonies transfected with the wild type HMG14a cDNA cloned into TFANEO in the sense and antisense orientation, and a 3'- untranslated region deletion mutant. Portions of the 3'untranslated region of the HMG14/17 proteins are very highly conserved suggesting they play an important role in gene function. It is possible that sequences within this region are important for mRNA processing, message stability, or lack thereof. Clones labeled 14-RS are HMGl4a cDNA with a 194 bp deletion in the conserved portion of the 3'-untranslated region. The results in figure 2 show the presence of two transcripts in the 110 Figure 1 - Map of expression vector TFANEO Expression. vector TFANEO is 6.8 Kb in length. Transcriptional controls are located in the expression cassette consisting of two LTR sequences derived from the Schmidt-Rupin Avian Rous sarcoma retrovirus. HMG14/17 cDNAs were cloned into the single CQaI restriction site located between the LTRs. A neomycin gene driven by the chicken B- actin promoter allows for selection by G418 in tissue culture. The arrows indicate the direction of transcription. 111 no. mo omz 8.0 a l ( Figure 2 114 transfected.cell lines. Presumably; the lower band.represents endogenous quail HMG14 message and the upper band, the exogenously transfected mRNA. Ethidium bromide staining of the RNA gel shows all lanes contain equivalent amounts of good quality, undegraded rRNA. Clones 14-RS-6 and 148-1 and 148-2 express the highest amount of exogenous mRNA. The contrOl cell lines QT6 and VC3, representing untransfected QT6 cells and QT6 cells transfected with the TFANEO vector only, show barely detectable levels of quail HMG14 specific mRNA. This suggests that quail HMG14 is expressed at very low levels in these cells or that the endogenous quail HMG14 message is sufficiently different in sequence from the chicken HMG14a so as to not hybridize efficiently with the chicken HMG14a cDNA probe used in these experiments. Most of the 14S and 14AS single colonies screened did not show elevated levels of exogenous HMG14a mRNA. Approximately 15 additional wild type and deletion mutant HMG14a clones were screened for mRNA expression in this manner (data not shown). The 3 overexpressing clones shown in figure 2 remain the highest expressors of exogenous chicken HMG14a mRNA seen in the cells tested. Interestingly, the 3'untranslated deletion mutant 14-RS-6 showed the highest levels of expression. This suggests that the deleted region may be involved in the regulation of mRNA expression levels. Figure 3 shows a similar experiment screening mRNA 115 Figure 3 - Nbrthern analysis of total RNA isolated from QT6 cells transfected with ENG17 cDNA constructs Total RNA was isolated from stable cell lines transfected with TFANEO containing various chicken HMG17 cDNA clones. 30 ug of total RNA was loaded per lane. The blot was probed with full length HMG17 cDNA. Two transcripts are seen, the upper band is presumably exogenous HMG17 mRNA and the lower, the endogenous quail HMG17. It is possible that the lower band represents an. additional alternatively' polyadenylated exogenous transcript. Transcript sizes were estimated by comparison to a simultaneously run DNA standard. QT6 is the untransfected.parental cell line. VC3 is QT6 transfected with TFANEO alone. 178 and 17AS are single colonies derived from stable transfections of TFANEO containing wild type chicken HMG17 cDNA.cloned.in.the sense and.antisense orientation. 17-R are single colonies from transfections of TFANEO containing HMG17 cDNA 3'-untranslated region deletions. 116 mummnbawzm mimmubawzm Nummnhawzm athIPHOZm Numdbawzm Htmflbawzm NaumbHUSm Hatmbawzm OH-MbHUZm mummawzm mummawzm b1mbawzm mumbawzm wumbfiwzm Mumbawzm Numbawzm HumbHOEm mU> who ~1500 nt t n O 0 3 1 ~ .- 21‘: i! . “H- , AI 0 Figure 3 117 expression levels in quail cells stably transfected with chicken HMG17 cDNA clones. 178 1-12 and 17AS 1,2 represent single colonies tranfected with.wild type HMG17 cDNAs cloned into expression vector TFANEO in the sense and antisense orientation. 17-RS 1,2 are HMG17 cDNA deletion mutants with a 178 bp deletion in the conserved portion of the 3'untranslated region. QT6 and VC3 are the control cell lines described above. As with the HMG14a transfected cell lines, most of the single colonies do not show'overexpression of the exogenous message. Clones 178 3,7 and 10 were the highest overexpressers seen in all cell lines screened, including an additional 10 clones tested but not shown. The fact that many stable colonies did not show overexpression of the exogenously transfected cDNAs may be explained in several ways. Very high levels of HMG14/17 mRNA may be lethal or deleterious to the QT6 cells, thereby naturally selecting only those clones which express low or moderate HMGl4/17 mRNA levels. Alternatively, some of the integrated vectors may have the HMG14/17 cDNA disrupted in the quail cell genome. Most likely, we believe that the amount of total HMG14/17 mRNA present in the cell is regulated at the transcriptional level by the cell, by sequences present in the 3'-untranslated region or by rapid degradation of excessively transcribed HMG14/17 mRNA. 118 Western analysis of ENG14/17 transfected cell lines. The amount of mRNA transcribed from.a particular gene does not.neccessarily reflect the amount of protein.translated and present in the cell. To address the question of whether the cell lines shown to overexpress HMG14 or HMG17 mRNA also overexpress the HMG protein, western analysis was performed with a variety of cell lines. HMG proteins were isolated from the various cell lines by a 5% perchloric acid (PCA) extraction of the nuclei. This isolation procedure has been shown to selectively extract the HMG proteins and histone H1. An antibody that recognizes both HMG14 and HMG17 was used to detect these proteins on the blots (66). The antibody was kindly provided by Michael Bustin. The antisera was elicited in rabbits using a synthetic peptide corresponding to the highly conserved 30 amino acid stretch of the human HMG17 DNA binding domain. This antibody has been shown to recognize both HMG14 and HMG17 specifically, with a slightly higher affinity for HMG17. Figure 4 shows a western blot of protein samples from several of the stable cell lines described above. In all lanes bands corresponding to the HMG14 and.HMG17 proteins are present and run at the appropriate mobility. To control for the total amount of protein loaded per lane, the gels were stained with coomassie blue and the lanes were normalized to 119 Figure 4 - Western blot of QT6 transfected cell lines HMG14/17 proteins were selectively isolated from cells with a 5% PCA extraction procedure. Approximately 1 ug of protein was loaded per lane. Proteins were run on 15% SDS-PAGE gels and electroblotted to PVDF membrane. HMG14/17 proteins were detected with an antibody specific to the conserved DNA binding domain of these proteins. The upper band is HMG14 and the lower band is HMG17. QT6 is the untransfected parental cell line. 14-RS-6,8 are two single colonies derived from transfection.with HMG14a cDNA 3'-untranslated region deletion mutants. 17S-7,10 are two single colonies derived from transfection of wild type HMG17 cDNA. Clones 178-7 and 17S-10 show 5 and 3 times overexpression of HMG17 as compared to the control cell lines. ~15 kd ~12 kd H (D Ad H (U E QT6 120 HMG14-RS-6 Figure 4 HMG17S-10 121 the slower migrating histone H1 band. The HMG14 / 17 bands seen on the radiographs were optically quantitated using AMBIS imaging software. The QT6 control cell lane in figure 4 clearly shows that more HMG14 is expressed in these cells than HMG17. We see this result consistently with our western blots, however published reports from M.Bustin et al. have shown that in human cell lines, HMG17 is expressed at slighlty higher levels than HMG14 (65). Cell line 14-RS-6, shown to express Ihigh levels of HMG14 specific mRNA.does not appear to express elevated levels of HMG14 protein” The ratio of HMG14/HMG17 is similar in QT6, 14-RS-6 and 14-RS-8, which does not show any elevated mRNA expression. This suggests that the excess HMG14 specific mRNA expressed in cell line 14-RS-6 is not efficiently translated or the translation product is unstable. Cell lines 178-7 and 17S-10, which were shown to overexpress HMG17 specific mRNA clearly show significantly higher levels of HMG17 protein as compared to the control cells. The level of total HMG17 expressed in these cells is consistently 3-5 times that seen in the control QT6 cells. Interestingly, 17S-10 displayed slightly higher levels of HMG17 mRNA than 17S-7 yet 17S-7 expresses slightly higher levels of HMG17 protein. In these cell lines it appears that this overexpression of HMG17 protein may be at the expense of HMG14. The amount of total HMG14 + HMG17 protein, normalized to histone H1, is roughly equal in the HMG17 overexpressing lines and the control 122 cells. This suggests that the cell may be regulating the total amount of HMG14/17 protein allowed in the cell. Figures 5-7 show representative western blots from a variety of stable cell lines transfected with wild type or mutant HMG14/17 cDNAs. The wild type and deletion mutant clones have been described above. 173-6 and 14A-1 are site specific mutants in the DNA binding domains of the respective HMG cDNAs. These mutant clones are described in detail in chapter III. Figure 5 shows clones 148-2 and 14S-3 do not overexpress HMG14 protein. 14S-3 was shown in northern.blots to express elevated levels of HMG14 specific mRNA while clone 14S-2 did.not. This again illustrates the lack of correlation with HMG14 clones between mRNA and protein expression levels. Figures 5 and 6 show protein expression screening from a total of 20 single colonies derived from a stable transfection of mutant 17B-6. There are no obvious overexpressers of HMG17 protein. Figure 6 also confirms the overexpression of HMG17 protein in clones 178-7 and 17S-10. Figure 7 illustrates HMG14/17 protein levels seen in 7 clones derived from.mutant 14A-1. None of these clones express excess amounts of HMG14. In addition to the samples shown in western blots shown here, another 29 clones were screened in this manner for HMG14/17 protein expression levels. These include clones derived from site-specific mutants 14A-1, 14A-4, 14C-1, 17B-6 and wild type cDNAs. Of 123 Figure 5 - Western blot of QT6 transfected cell lines HMG14/17 proteins were isolated from stable cells lines transfected with various HMG14/17 cDNA clones. Approximately 1 ug of protein.was loaded.per lane on a 15% SDS-PAGE gel. The proteins were detected with an antibody specific to the conserved.DNA.binding domain of these proteins. The upper band is HMG14 and the lower is HMG17. QT6 is the untransfected parental cell line. 14S-2,3 are two single colonies derived from transfection of wild type HMG14a cDNA. 17B6 clones are single colonies derived from transfection of a HMG17 cDNA containing the B6 site-specific mutation in the conserved.DNA binding domain. There are no apparent overexpressers of HMG14 or HMG17 in this panel. 124 mMImmbHOEm mMImmbHOEm VMnmmbHUSm mmimmbawzm OMummhawzm mmummbawzm mmummbawzm maummbflwzm mummbHUEm mimmawzm Nummaozm who HMG14 HMG17 .1". m—nA-I—llI-IIFd-l «II-o-” " “--u-w Figure 5 125 Figure 6 - Western blot of QT6 transfected cell lines HMG14/17 proteins were isolated from stable cell lines transfected with chicken HMG17 cDNA clones. Approximately 1 ug of protein was loaded per lane on a 15% SDS-PAGE gel. The proteins were detected with an antibody specific to the conserved DNA binding domain of these proteins . The upper band is HMG14 and the lower band.is HMG17. QT6 is the untransfected parental cell line. 178-7,10 are two single colonies derived from transfection of a wild type HMG17 cDNA. 17B6 clones are single colonies derived from a transfection of HMG17 cDNA containing the B6 site-specific mutation in the conserved DNA binding domain. Clones 17S-7 shows overexpression of HMG17 5 times and 178-10 2 times that of the control cells. None of the HMG17B6 mutants show overexpression of HMG17. 126 mmnmmbawzm Hmommbawzm hmummbawzm mmummbaozm mmnmmbflwzm mmnmmbawzm vaummbawzm vummbawzm mimmbHOEm Nummbflwzm HummbHOEm OHumbHUZE bumbaozm GEO .jfl' HMG14 .‘W HMG17 CID-C-D‘l.”"-i.-_ ._. Ls 1.54-DV‘.' «a - .n. ‘u- 0!. ‘Illiem. anqfllhdlo 9" - “+- ~1- Figure 6 127 Figure 7 - Western blot of QT6 transfected cell lines HMG14/17 proteins were isolated from stable cell lines transfected with chicken HMG14a cDNA clones. Approximately 1 ug of protein was loaded per lane on a 15% SDS-PAGE gel. The proteins were detected with an antibody specific to the conserved DNA binding domain of these proteins. The upper band is HMG14 and the lower band is HMG17. The bands visible below HMG17 are probably HMG14/17 degredation products. The band visible between HMG14 and HMG17 may be HMG14b which is expressed at low levels in avian cells. QT6 is the untransfected parental cell line. Clones 14-RS-6,8 are two single colonies derived from a transfection of HMG14a cDNA with.a:3'-untranslated.region.deletion. 14A1 clones are single colonies derived from a transfection of HMG14a cDNA containing the A1 site-specific mutation in the conserved DNA binding domain. There are no apparent HMG14 overexpressers in this panel. ' 128 HHquwawzm muadvawzm mtadmflozm bufiflwdwzm wIH4VHUZZ mufldvfiozm vtfldeOEm mummuvHUEm muwm|¢awzm who Figure 7 129 all the clones screened, clones 178-7 and 17S-10 were the only ones that showed significantly overexpressed levels of HMG14 or HMG17 protein. Growth.rate and morphology of HHGI4/17 transfected.cell lines To examine the possible morphological effects on the QT6 cells due to overexpression of HMG14 or HMG17 mRNA or protein, various stable cell lines were microscopically photographed and/or examined visually. There was no apparent gross morphological differences seen between the cells lines examined and the QT6 parental cells. It is possible that overexpression of the HMG14 or HMG17 mRNA or protein will affect the growth rate of the cells. Growth curves were done for several cell lines including clones 14-RS—6 and 178-7. As seen in figure 8, 14-RS-6 grew at a significantly slower rate than the control cells. The doubling time was approximately 48 hours for the 14-RS-6 clone compared to less than.24 hours for QT6. As shown above, clone 14-RS-6 expresses high levels of HMG14a mRNA.but not HMG14a.protein. The decrease in.growth rate seen with this clone may be influenced by this overexpression but it also could be a chance effect of the intregration site of the exogenous DNA in the host cell chromosome. Clone 17S-7, which overexpresses HMG17, showed a doubling time similar to QT6. 130 Figure 8 - Growth curve for control and transfected QT6 cells. A.6 day growth curve for untransfected.QT6 cells and 3 cell lines transfected with ITANEO driven constructs. All time points were done in duplicate. Plates were seeded with equal numbers of cells on day 0 and duplicate plates were counted for cell number at each time point. QT6 is the untransfected parental cell line. VC3 (vector control 3) is QT6 transfected with TFANEO only. 14-RS-6 is a transfected cell line with the HMG14a cDNA containing a 3' -untranslated region deletion. 17S- ? is a cell line transfected with the HMG17 cDNA in the sense direction. 131 CeHruunberx 10000 woo moo 000 #00 woo woo Loo 44.0 new zom - <: ( coocmcco\ coeoeoto >20 59.9 :3 m-®ro:aa 86.5.38 .2050 to $0833 >20 Essa toeotodo 02.3. 29:2: Figure 9A 138 :65: :65: _ _ 6.8 8 an: > 2.9 8388 o e >20 :29: 9.53 for: 2.8 Eoow _ L 5.8 so. for: 9: .89 2. 8.0.5.5 Figure 9B 139 transfections). The various HMG14a clones were transiently and stably transfected intotQT6 cells as described in earlier in this chapter. Protein.was isolated from.the transfectants and western analysis was used to determine the amount of HMG14a expressed in these clones. Protein isolation, gel conditions, loading controls, and antibody detection were done as previously described in this chapter. Western analysis of HMGl4a transfected cell lines Initial transient transfections were done to establish whether the tetracycline regulatable vector system would express HMG14a in our QT6 tissue culture system. Figure 10 shows a western analysis of two such transfection experiments. Lane 1 is a mock transfection with no DNA, and lane 2 is a co-transfection of pUHDIS-l and pUHD10-3 containing the HMG14a coding region only. The transfections were done in the absence of tetracycline which should allow maximal expression of the HMG clone. The HMG14 band in lane 2 is at least 3.5 times greater than the corresponding endogenous band in the mock transfection. This suggests that the HMG14a clone is being expressed at easily detectable levels using the 2 vector expression system in our QT6 cell line. This is despite the fact that in this, as in any other transient system, many cells may not have been successfully transfected with either or both DNAs necessary for 140 Figure 10 - Western analysis of protein isolated from cell lines transiently transfected with the coding region of HMG14a HMG14/17 proteins were isolated from cells 24 hours post- transfection with a 5% PCA extraction as described in chapter II. Approximately 1 ug of protein.was loaded.per lane on a 15% SDS—PAGE gel. The HMG proteins were detected with an antibody specific to the conserved DNA domain of these proteins as described in chapter II. The upper band is HMG14 and the lower is HMGl7. Transfections were done in the absence of tetracycline which should allow maximal transcription of the HMG14-containing clone. Lane 1 represents the mock-transfected QT6 and lane 2 represents QT6 transfected with 10 ug of pUHD10—3/HMG14a coding region and 10 ug of pUHD15-1. In lane 2, HMG14 is being expressed at least 3.5 times that seen in the control-cells. The levels of expression were optically quantitated using AMBIS imaging software. The level of HMGl4 protein expression determined in lane 2 is a minimal value due to the intensity of the band which is out of the linear analyzing range of the AMBIS software. 141 Lane 1 Lane 2 "' HMG17 Figure 10 142 regulatable HMG14 expression. Single and pooled colonies were isolated from stable transfections of HMG14a wild type and coding region only clones in pUHD10-3 co-transfected with pUHD15-1 and TFANEO. Figures 11 and 12 show western.blots of HMG protein isolated from various cell lines derived from these transfections. The HMG protein samples were isolated with a 5% PCA extract of total cells. Single colonies were grown in the presence of tetracycline and split to 4 plates each. The media was changed to tetracycline free media and protein was isolated from each clone at 0, 6, 12 and 24 hours. The time course was done to determine the kinetics of protein expression over the first 24 hours of promoter induction. Surprisingly, there was no increase in HMG14 expression seen in any of the single colonies shown as well as in 22 other single colonies and 2 pooled colonies not shown. Protein isolated from clones after 6 and 14 days in tetracycline free media showed similar results (data not shown). The lack of exogenous HMG14a overexpression seen with the stable cell lines tested could.be due to some type of cellular regulation of the HMG protein levels, or the absence of one or more of the 3 vectors used in the transfection. The stable transfectants were isolated in the presence of G418 in the tissue culture media; thus, the vector TFANEO must be present in all the stable cell lines. To determine 143 Figure 11 - Western analysis of protein isolated from cell lines stably transfected with the coding region of HMGl4a HMG14/17 proteins were isolated from cells with a 5% PCA extraction as described in chapter II. Approximately 0.5 ug protein was loaded per lane on a 15% SDS-PAGE gel. The HMG proteins were detected with an antibody specific to the conserved DNA domain of these proteins. Stable colonies were grown in the presence of tetracycline and.protein.samples were isolated 0, 6, 12, and 24 hours after the media was changed to tetracycline free. Clones HMG14H-6 and HMG14H-7 are two single colonies isolated from.a stable co-transfection of the coding region of the HMG14a cDNA cloned into pUHD10-3 (H signifies the HincII restriction fragment of the HMG14a cDNA which contains the coding region) along with pUHD15-1 and TFANEO. No perceptible increase in HMG14 expression was seen over the first 24 hours in either clone. 144 HMG14H-7 HMG14H-6 HQ mm 2 2 Bo use MS mm ME NH H: m use HMG14 HMG17 Figure 11 145 Figure 12 - Western analysis of protein isolated from cell lines stably transfected with the HMGl4a cDNA HMG14/17 proteins were isolated and detected as described in figure 11. Protein samples were taken at 0, 6, 12, and 24 hours after thewmedia.was changed.to tetracycline free. Clones HMG14E—7, HMG14E-8, HMG14E-9 and HMG14E-10 are single colonies isolated from.a stable transfection of the full length HMG14a cDNA cloned into pUHD10-3 (E signifies the EcoRI restriction fragment containing the full length HMG14a cDNA) along with pUHD15-1 and TFANEO. No perceptible increase in HMG14 expression was seen over the first 24 hours in any of these clones. 146 HMG14E-8 HMG14E—7 .3 E E 3 E m E o E am E S E 6 Bo HMG14 HMG17 HMG14E-10 HMG14E-9 E em E S E m .3 o E «m E S E m Bo HMG14 HMG17 ’ as! a. Figure 12 147 whether the VP16-TetR fusion activator protein was being expressed from vector pUHD15-1 in these cell lines, pUHD10-3 containing a luciferase gene cloned in place of the HMG14a clone was transiently transfected into pools of the stable cell lines (73). Expression of the fusion activator protein should lead to high levels of luciferase activity in the absence of tetracycline. A portion of the transfected cells were analyzed for luciferase activity and protein was isolated from the remaining cells and analyzed by western blot for HMG14a expression. The results of this experiment are shown in figure 13. As can.be seen from.the western.blot, there is no apparent increase in HMG14a expression in any of the pools tested. However, the data from the luciferase assay clearly show that the luciferase gene is induced in tetracycline free media indicating that the activator protein is, in fact, expressed and is functional. Thus, the tetracycline regulatable system.appears to be operative in the clones tested, yet there is no apparent increased expression of the HMG14a protein in these clones. Discussion It is possible that the cell will tolerate only a limited amount of HMG protein expressed in the cell. Possible mechanisms for this regulation have been discussed in the previous discussion section. The use of a regulatable expression.vector to express exogenous HMG14a may be one way 148 Figure 13 - Western analysis and luciferase assay of protein isolated from.stably transfected clones transiently transfected with pUHD10-3 containing the luciferase gene. A. Western analysis HMG14/17 protein was isolated and detected as previously described from.pooled colonies of stable transfectants. Tetracycline (4 ug/ml) was present or absent in the media as indicated by +/- tet. HMG14E represents the full length HMG14a cDNA cloned into pUHD10-3. HMG14H represents the coding region only portion of the HMG14a cDNA cloned into pUHD10-3. P indicates pooled stable transfectants.The pools for each construct were transiently transfected with the luciferase gene cloned into pUHD10-3. Mock represents the stable colonies mock transfected with no luciferase DNA. There was no increase in HMG14 expression in any of the clones tested upon tetracycline removal. B. Luciferase assay Samples from each of the transient transfections were analyzed for luciferase activity. Elevated levels of luciferase activity were seen in transfections carried out in the absence of tetracycline. 149 p.11: am A Te 1: + + I I t I 2 2 z o. o. o. o. o. o. :1: :1: :1: :r: m m V V‘ S” Q‘ <1' Sl‘ ...'. . v-i r-1 v-l v—i r-i r—i :.:..: C29 (.9 g [23 (S9 (29 r E m :r: :r: :1: g canccslm4 EC _ __ _ A- HMG l 7 ,fl 5% B. Clone Tet Mock Luciferase activity 3:55 HMG14HP + Y 0 - 00 HMG14HP + N 293.60 HMG14HP - Y 0.00 HMG14HP - N 1767.00 HMG14EP + Y 0.00 HMG14EP + N 500.10 HMG14EP - Y 3.16 HMG14EP - N 2378.00 Figure 13 150 to overcome potential feedback regulation of the HMG protein levels. A sharp induction of expression may transiently elevate the amount of HMthrotein.before the cell can adjust to the normal levels. However, our experiments suggest that either this is not the case or if feedback regulation of excess HMG expression exists, it is most likely a rapid process. Bujard et al. showed that expression of luciferase from vector pUHD10-3 in the presence of pUHD15-1 in tetracycline free media reached maximum levels at 12-24 hours (73). Our time course experiments showed that HMG14a expressed from.pUHD10-3 did not show elevated levels of expression from 0-24 hours nor after 6-14 days. It remains a formal possibility that none of the pUHD10-3/HMG14a DNA was taken up by the cells along with pUHD15-1 and TFANEO during the stable transfections. However, a number of pooled colonies were tested for protein expression and the chance that this particular DNA.would be preferentially excluded in all the transfected cells seems highly unlikely. It is also possible that all the transfecting pUHD10-3/HMG14a DNA copies disrupted in the HMG14a cDNA sequence (or in nearby flanking sequences required for expression) during integration into the QT6 genome. However, this again seems rather unlikely. Due to the fact that we can not distinguish between the endogenous and exogenously expressed HMG14 protein with the antibody used in these experiments, there remains the 151 possibility that the protein detected is in fact the exogenous chicken HMG14a and that the endogenous quail protein is being repressed. Alternatively, our HMG14a cDNA clones may contain unknown elements which make them unrecognizable or inefficient in translation. To address these last possibilities, HMG14a cDNA clones were prepared which contain an additional 24 hp at the 3' end of the coding region that, when translated, code for eight amino acid residues recognized with high specificity by a commercial antibody. These clones and their expression are discussed in- chapter V. Chapter V Expression of HMG14aFLAG in tissue culture cells 152 153 Chapter V INTRODUCTION Detection of total HMG14 and HMG17 protein.expressed.in.our stably transfected.tissue culture cell lines with an antibody specific to the DNA.binding domain.of these proteins does not allow us to distinguish between the endogenous quail HMG14 and the exogenously transfected chicken.HMG14a. Our previous results show that very few transfected clones express elevated amounts of the transfected protein and.we cannot be sure that this additional expression is, in fact, due to the exogenously transfected.HMG cDNA, or rather, due to aberrant endogenous HMG expression. While it is possible that the inability to isolate numerous highly expressing HMG14/17 clones is due to an unknown cellular regulatory mechanism which limits the amount of HMG14/17 protein allowed in the cell, we cannot be sure the exogenous HMG14/17 clones are translated efficiently or remain stable once translated. To date, none of the point mutations made in HMG14 or HMG17 has resulted in an obvious change in electrophoretic mobility. To address these concerns, we prepared.HMG14a.wild.type and mutant cDNA.clones that contain an additional 24 bp at the 3' end of the coding region such that an additional 8 amino acids are fused to the C-terminus of the translated protein. These eight amino acid residues are recognized with high affinity by a commercially available antibody, FLAG M2 (80). 154 This allows us to specifically detect the exogenously transfected HMG14 and therefore address the issues raised above. The C-terminus of the HMG14a protein is highly acidic. It is possible that the residues located here (or at least their negative charge) directly participate in the protein's ability to function in the cell. Though no function has been assigned to this region.of the protein, it is thought that it may be involved in binding to histones or to other HMG molecules on the nucleosomal surface. The 8 amino acid FLAG sequence added to the end of the coding region is also highly acidic and its overall charge properties are similar to the HMG14 C-terminus. It is hoped that the additional sequences will not disrupt the protein's ability to bind to the nucleosome. 155 RESULTS Transfection and expression system Wild type and mutant HMG14aFLAG clones were stably and transiently transfected into QT6 cells as previously described. The FLAG constructs were cloned into the expression vector pUHD10-3 and expression in culture was regulated by the presence or absence of tetracycline in the media as previously described in chapter IV. Protein isolation and western blotting were done as previously described except for the use of the anti-FLAG M2 antibody in addition to the anti-HMG antibody. Construction of the HMG14aFLAG cDNA clones HMG14a cDNA cloned into plasmid pCla12 (pC14cw) was used as a template in a PCR scheme to prepare the wild type HMG14aFLAG construct. The basic PCR scheme used was the same as for the site-specific mutagenesis described previously using complementary internal primers containing the additional FLAG sequences instead of primers containing single base mismatches. The internal FLAG primers used are listed in figure 1. The PCR scheme is illustrated in chapter III, figure 6. The two outside primers used, A' and D' are described in chapter III, figure 5. For production of the site-specific mutant HMG14aFLAG constructs, the wild type HMG14aFLAG in pCla12 was used as a template and the internal 156 Figure 1 - Oligonucleotide prrmers for PCR production of HMG14FLAG constructs Oligo name Wild type FLAGl-AS FLAGZ-S Murant 14C3-S 14C3-AS 14A2-S 14A2-AS 14Bl-S 14Bl-AS Sequence 5' - 3' CTTGTCATCGTCGTCCTTGTAGTCCTCGGACTTGGCTTCC ************************ GGACGACGATGACAAGTAATGTTAACCCTGCC **************** * = FLAG sequence CAAAGCTCAAAAAGGCAGC GCTGCC'I'I'I'I'I'GAGCT'I'I‘G GAGCCAAACAGAAGG cc'I-rc'rcrrrccc'rc GTCGGCCGGACTATCTGC GCAGATAGTCCGGCCGAC 157 mutagenic primers used are listed in figure 1. Portions of the coding regions of the resultant FLAG constructs were verified.by DNA sequence analysis (Michigan State University sequence facility). The wild type and mutant HMG14aFLAG cDNA fragments produced in the PCR scheme were digested with EcoRI (see chapter III, figure 6) and cloned into the EcoRI site in the pUHD10-3 polylinker for use in transfection experiments. Figure 2 shows an illustration of the complete HMG14aFLAG cDNA clone as well as the DNA and amino acid sequence of the C-terminal region of HMG14a with and without FLAG sequences. Transient expression of HMG14aFLAG constructs Initial transient transfections of the HMG14aFLAG in expression vector pUHDlO-3 were done to establish whether the protein would be expressed and could be detected with the anti-FLAG M2 antibody specific to the 8 amino acids added to the 3' end of the HMG14a coding region. 10 ug of the wild type HMG14aFLAG/pUHD10-3 DNA and 10 ug of vector pUHD15-1 (expressing the transcriptional transactivator protein) were transfected onto plates of QT6 cells at approximately 60% confluency. Duplicate transfections were performed in the presence or.absence of tetracycline (4.0 ug/ml) in the media. The HMG14aFLAG protein should be expressed in the absence of tetracycline. HMG protein was isolated with a 5% PCA 158 Figure 2 - Schematic of HMG14aFLAG construct and comparison of C-terminus of HMGl4a with and without FLAG sequences A. Schematic illustration of HMG14a wild type cDNA with fused FLAG sequences at the 3' end of the coding region. The DNA and amino acid sequence of FLAG is shown. B. Comparison of the C-terminal residues of HMG14a coding region with and without FLAG sequences added starting at residue #72 (out of 104 in HMG14a). Acidic and charged residues are labeled. (* = acidic residue, + = basic residue) Amino acid abbreviations are listed in appendix. 159 axis: r H“ A ' um FHCHaFLAC cDNA :23: no m 1 coding region FLAG / 990 MW '/////////////////////////M/ ///////,: //7//////////////7/ ////////.////// W/Mfl/fl :::';:,-' SUT / \ 3‘” 15:13 . GACTACAAGGACGACOATGACAAG asi‘ AsoTurLusAspAspAsnAsoLus B . HMGl4a EDAKEENHSENGDTKTNEAPAAEASDDKEAKSE EMG14 aFLAG .AKEEN'HS ENGDTKTNEAPAAEASDDKEAKS EDYKDDDDK ** +** + * * + * * **+* + ** +****+ Figure 2 160 extraction at 48 hours and the protein was electrophoresed on 15% or 18% SDS-PAGE gels. Detection of FLAG containing proteins was done using anti-FLAG M2 antibody purchased from Kodak-IBI . This antibody recognizes the FLAG protein sequence with high affinity when it is located at the C-terminus of a polypeptide (for N-terminal FLAG constructs, another antibody, anti-FLAG M1 can be used) . Figure 3 shows a western blot of a transient transfection.with the cells grown in the presence or absence of tetracycline. As expected, in the transfection containing tetracycline in the tissue culture media, no HMG14aFLAG was detected (lane 2). In the transfection containing no tetracycline in the media, the HMG14aFLAG protein is clearly expressed and is detected as a single band migrating at the expected mobility of approximately 14 kd (lane 2). This result demonstrates that the HMG14aFLAG clone is expressed and translated efficiently. Because the cDNA clone is the same as used in all previous experiments, with the exception of the extra FLAG sequences at the 3' end, this result demonstrates that the cDNA clone is able to be expressed and translated effectively and that the lack of overexpression seen previously is not due to any problems in the construction of the cDNA clones or in their ability to be translated. In addition, these results demonstrate that the regulated tetracycline expression system is functional and that expression of the exogenous gene is 161 Figure 3 - Detection of HMGl4aFLAG by Western blot QT6 cells were transiently transfected with plasmid constructs as listed in the figure. Cultures were grown.in the presence (4.0 ug/ml) or absence of tetracycline in the medial Protein was isolated at 24 hours post-transfection with a 5% PCA. extraction. as described. in chapter II. Protein ‘was electrophoresed on 18% SDS-PAGE gels, blotted and detected with anti-FLAGM2 antibody (80). HMG14aFLAG-protein is clearly’ detected in lane 3 while none of the control lanes showed evidence of expression. 162 pUHD10-3/ HMG14aFLAG-S - + + - — — _ _ _ pUHD10—3/ HMG14a-AS - - — _ - - + + _ pUHD10-3/ HMGl4a—S - — - - + + — _ .. pUHD15-l - + + + + + + + + Tet ‘ + - + — + — + _ Figure 3 163 tightly controlled. HMG proteins were isolated from.the transfected cells with a 5% PCA total cell extraction. For the exogenously transfected protein to exert any effects on the chromatin structure within the cell, it is necessary that these proteins be located in the nucleus and not solely in the cytOplasm. To demonstrate the cellular location of the transfected protein, transient transfections were performed as above with each plate being divided into nuclear, cytoplasmic and total cell fractions. In.addition to the wild type HMG14aFLAG, a site—specific mutant in HMG14a, HMG14aC3 -FLAG was also used in these transfections. The HMG14aC3 mutant contains a single base pair mutation at codon 35 in the cDNA sequence resulting in a proline to lysine change in the protein sequence. 5% PCA extractions were done on each fraction.and the isolated protein.was run.out on 18% SDS-PAGE gels for analysis. Figure 4 shows the results of the western.blots using the anti-FLAG M2 and anti-HMG antibodies to detect the protein. In figure 4A, the anti-FLAG M2 antibody detects HMG14aFLAG protein only in the nuclear and total fractions in.the absence of tetracycline. In figure 4B, the anti-HMG antibody detects total HMG14/17 in only the nuclear and.total fractions. These results therefore indicate that the wild type HMG14aFLAG and the mutant HMG14aC3FLAG are being expressed and the protein is localized to the nucleus. 164 Figure 4 - Detection of HMGl4a and EMGI4aFLAG proteins in cell fractions from transiently transfected cells A. QT6 cells were transiently transfected with plasmid constructs as listed in the figure. Cultures were grown in the presence (4.0 ug/ml) or absence of tetracycline in the media. Protein was isolated with a 5% PCA extraction from nuclear, cytoplasmic and total fractions of transfected cells at 24 hours post-transfection. The fractions were prepared as follows. Nuclei were isolated by cellular lysis as described in chapter II and the cell lysate was taken as the cytoplasmic fraction. Nuclear pellets were washed 3-5 times in lysis buffer to rid nuclei of lingering cytoplasmic proteins and washes were added back to cytoplasmic fractions. Nuclear intactness was determined by microscopy. Total fractions refer to a 5% PCA extraction of whole cells. Protein was loaded (approximately 10° cells/ lane) and electrophoresed on 18% SDS- PAGE gels. Gels were blotted and protein was detected with anti-FLAGM2 antibody. HMG14aFLAG wild type and mutant proteins were detected only in the nuclear and total cell fractions of the appropriate transfections. B. Western blots from figure 4A were stripped and reprobed with anti-HMG antibody. The lanes correspond to those in figure 4A. HMG14/17 was detected in all nuclear and cytoplasmic fractions. 165 pUHDlO-3/ HMG14aFLAG-S l I l + + + + + + pUHDlS-l + + + + + + + + + Cell fraction N C T N C T N C T Tet + + + + + + - - - 10 ll 12 13 14 15 16 17 18 pUHD10-3/ HMG14C3FLAG - — - + + + + + + pUHD15-l + + + + + + + + + Cell fraction N TEC + Figure 4A 166 10 ll 12 13 14 15 16 17 18 9. .. HMG14 F - n HMG17 Figure 4B 167 In addition, this demonstrates that there is no significant leakage from nuclei isolated by our standard procedure. Stable transfections of HMG14aFLAG constructs The HMG14aFLAG and HMG14aC3FLAG constructs were stably transfected into QT6 cells. Single colonies were grown up in the presence of tetracycline to repress expression of the FLAG constructs. Protein was then isolated from duplicate plates of cells after 48 hours of growth with or without tetracycline in the media. 25 single colonies of HMG14aFLAG and 8 colonies of HMG14aC3FLAG transfectants were tested by western analysis for expression of the respective FLAG containing proteins. Results of several of these experiments are shown in figures 5 and 6. As can be seen in the blots probed with anti-FLAG M2 antibody, there was no FLAG containing protein detected in any of the single colonies tested. In most lanes, two series of bands were detected at high mobilities inconsistant with the expected size of the FLAG proteins. These bands are present in the untransfected QT6 cell lines as well and therefore represent cross reactivity with unknown proteins present in the QT6 cell line. All the blots were reprobed with the anti-HMG antibody showing endogenous HMG14/17 protein.was detectable at normal levels as demonstrated in figure 6. The results with the transient transfections demonstrate 168 Figure 5 - Western blot of protein isolated from stably transfected cell lines Single colonies were isolated from. QT6 cells stably transfected with TFANEO, pUHD15-1 and pUHD10-3 containing HMG14FLAG (designated F) or HMG14a cDNA (designated E) . Cells were grown in the presence or absence of tetracycline for 24 hours and protein was isolated, electrophoresed, and blotted as previously described. Blots were probed with anti-FLAGM2 antibody. There was no detection of HMG14FLAG in any of the cell lines tested. Several lOWWmobility'bands'wererdetected.in all lanes that do not correspond to known proteins. M refers to a FLAG control protein purchased from Kodak-IBI (80) and serves as a positive control for the anti-FLAGMZ antibody. 169 Clone # M F41 E4 F43 F45 F19 Tet + - + - + - + - + "we. «I»,- e: . -.Hlihn - ea agij' .5 Clone # M F19 F34 F14 F19 F42 Tet + - + - + - + — I. --- -- - .. 1...... tr {’01 . and. a! Li 7. Figure 5 170 Figure 6 - Western blot of protein isolated from.stably A. transfected cell lines Protein was isolated from QT6 cells stably transfected with plasmids TFANEO, pUHD15-1, and HMG14C3FLAG/pUHD10-3. Cells were grown in the presence or absence of tetracycline and the protein was isolated, electrophoresed, and blotted as previously described. The protein was detected with anti-FLAGMZ antibody; There was no HMG14FLAG detected in.any of the cell lines tested. Lane M refers to a FLAG control protein purchased from Kodak-IBI (80) and serves as a positive control for the anti-FLAGM2 antibody. QT6 refers to untransfected control cell line. P refers to a pooled colony derived from the transfection and 1,2, and 6 refer to individual single colonies. Duplicate gel run simultaneously with gel in 6A and detected with anti-HMG antibody. Normal amounts of HMG14/17 are seen in all lanes. 171 Lane P l l 2 2 6 Tet - + — + - + - Lane QT6 P P 1 l 2 2 6 6 Tet - + — + — + - + - t... .. .- .. ...... HMG14 HMG17 Figure 6 172 that the constructs used in these experiments are able to be expressed and detected by our antibodies. The inability to isolate stable colonies that express high levels of FLAG containing protein could be explained in several ways. The cell may only be able to tolerate a limited amount of HMG14 protein such that additionally expressed HMG14 is rapidly degraded or there may be limited sites available on the nucleosomes for HMG14 and additional non-bound protein is rapidly degraded. If this is the case, the exogenous chicken derived fusion protein may be preferentially degraded over the endogenous quail protein. Alternatively, perphaps there is a copy number effect, such that in the copy number may be low to moderate in stables, but very high in transients. Finally, since the plasmid containing the HMG14FLAG constructs is unselected in the cell, over the time required to grow the single colonies to to the point where there are enough cells from which to isolate protein, about 3-4 weeks, the plasmid could.be lost or may become methylated or otherwise modified such that expression is severly limited or repressed altogether. PCR detection of plasmid sequences in transfected cell lines To determine whether the transfected HMGl4aFLAG plasmid sequences are present and uninterrupted in the stably transfected cell lines, PCR was used to detect diagnostic 173 Figure 7 - PCR strategy for detecting HMGl4aFLAG/pUED10-3 sequences in stably transfected cell lines A. Schematic of plasmid HMG14aFLAG/pUHD10-3 integrated in the cell genome. pUHD10-3 promoter (460 bp) contains the hCMV minimal promoter with heptemerized Tet operator sequences fused upstream.(73). Genomic DNA was isolated from stable cell lines, digested with HindIII, and amplified.with the primer pairs shown. Primer 1 (sense) binds to sequences within the Tet operator and was paired with.primers 2,3 and 4 (anti-sense) that bind to sequences within the HMG14a cDNA coding region. B. Sequences of primers used to amplify portions of HMG14aFLAG/pUHD10-3. 174 8 82 Y _ 3 a: m- _ as com a. _ and toe—ta y... AI _ 2:05 38.6 :82! «So 054.292 856.. 8.08.3. 085.63 928.30 5 n 93“ 9.68 S h 9293: 88.38 25.3 \ v::_::dm __ _ wNNNNNNNNm :55: zoom :83 .99. V8. 98— 3v _ Figure 7A 175 Primer Sequence 5' - 3' 1 (Tet operator) GTGAAAGTCGAGTTTACCACTCC 2 (14C3-AS) GCTGCCTTTTTGAGCTTTG 3 (14FLAG-AS) CTTGTCATCGTCGTCCTTGTAGTCCTCGGACTTGGCTTCC 4 (14SEQPR2) CCAAAACTAATGAGGCACCAGCTGC Figure 7B 176 fragments in genomic DNA isolated from several of the colonies tested. The PCR strategy and primers used are illustrated in figure 7. Primer pairs used included a sense primer hybridizing to Tet operator sequences in the 5' end of the HMG14aFLAG/pUHD10-3 and several antisense primers hybridizing to sequences internal in the HMG14a coding region. Primer pairs were designed to test for the intactnesss of the promoter region and most of the FLAG cDNA sequence. Amplification products were electrophoresed on agarose gels and primer pair #3 generated DNA fragments corresponding to the expected size (770 bp, see figure 7A) were in 4 of the 6 clones tested. These results indicated that at least some of the stable colonies contained the HMG14aFLAG plasmid intact within the quail genome (presumably at more than one copy per haploid genome). Western analysis of HMGl4aFLAG stable clones One of the previously unscreened single colonies which tested positive using PCR for the presence of the HMG14FLAG containing plasmid, HMG14C3FLAG-18, was analyzed for protein expression along with QT6 and an untested pooled colony, HMG14C3P. Cells were grown and protein was isolated as previously described. Protein samples were electrophoresed on 18% SDS-PAGE gels, blotted and probed with anti-HMG and anti-FLAGMZ antibodies. Figure 8 shows the results of this experiment. There was no HMG14FLAG protein detected in either 177 Figure 8 - Western blots of protein isolated from stably A. transfected cell lines Protein was isolated from the previously untested cell line HMG14C3FLAG-18 and from control QT6 cells. Cells were grown in the presence or absence of tetracycline and the protein was isolated and blotted as previously described. Protein was detected with anti-FLAG M2 antibody. Lane M refers to a FLAG control protein purchased from Kodak-IBI (80) and serves as a positive control for the anti-FLAG M2 antibody. There was no HMG14aFLAG protein detected. . Duplicate gel to that in A above, using anti-HMG antibody. Normal levels of HMG14/17 are seen in all lanes. Tet M QT6 18 178 18 Figure 8 HMG14 HMG17 H .1! 179 of the cell lines tested. The anti-HMG antibody detected normal amounts of HMG14 and HMG17, while the anti-FLAGMZ antibody detected only the previously seen high molecular weight bands. The repeated inability to isolate stable colonies that express significant levels of exogenously transfected HMG14a protein is therefore not due to loss or disruption of the HMG14—containing plasmid in the stable lines. Discussion The inability to isolate stably transfected.QT6 cells that express high levels of the exogenous HMG14FLAG proteins from plasmid pUHD10-3 is consistent with our results using plasmid vector TFANEO. Previous experiments indicate that the HMG14aFLAG constructs are reproducibly expressed and the protein is detectable in transient transfection experiments. In at least 60% of our stable cell lines, the plasmid sequences are present and uninterrupted in the QT6 genome. In transient and stable cell lines the tetracycline regulated expression system has been shown to tightly control expression of HMG14a, HMG14aFLAG and luciferase gene constructs. The abSence of detectable HMG14aFLAG in the stable cell lines tested is therefore most likely due to one or more cellular regulatory mechanisms functioning to repress expression or eliminate excess HMG14 protein. The question of cellular regulation of HMG protein levels 180 raises several possibilities. It may be that only HMG14 bound to chromatin is stable. It is known that the HMG14 proteins are found bound only to a subset of nucleosomes in-vivo. If the limited binding sites are occupied by endogenous quail protein, and the exogenous chicken HMG14aFLAG does not compete efficiently for these sites, then the cell may rapidly degrade the unbound.protein. Although.HMG14 proteins show a high level of evolutionary similarity, the chicken HMG14a proteins may be sufficiently different from the (as yet uncloned) quail HMG14, particularily with the added FLAG tail, as to be relatively unstable or unable to efficiently compete for the limited binding sites available on chromatin. It may also be that the chicken HMG14a mRNA levels are feedback or otherwise regulated by the quail cell. The transfected.HMGl4a.proteins were reproducibly detected in transient transfections, yet were not detectable in cell lines derived from stable transfections. The expression of stably integrated plasmids may be repressed by a number of mechanisms. Expression can be influenced by the location of integration in the genome, the presence or absence of regulatory elements controlling gene expression over large distances, or presence of endogenous repressor activity (84) . Alternatively, the exogenous HMG genes contain high G + C levels and once integrated, could be silenced.by deenovo CpG methylation (85) . Finally, the lack of detectable protein may 181 be due to a copy number difference. The endogenous quail HMG14a is known to be present at 2 copies per cell genome. The HMGl4aFLAG is clearly detectable in transients where the copy number is probably greater than 100 copies per cell. In stable transfectants, the copy number is presumably much lower, at 1-10 copies per cell genome, and thus may account for a greatly reduced or undetectable expression level. Chapter VI Chromatin structural studies of cell lines ectopically expressing chicken HMG14a and HMG17 182 183 Chapter VI INTRODUCTION The HMG14 and HMG17 chromatin binding proteins have been highly conserved throughout evolution suggesting they provide a vital cellular function. The location and specificity of their binding suggests a role in the regulation of gene expression. These proteins bind to the subset of chromatin known to be in an active or potentially active transcriptional state. The amount of HMG14 and HMG17 present in the cell is regulated and limits binding to 1 out of every 10 nucleosomes, with up to two molecules bound per nucleosome. Despite a great deal of research characterizing the specific molecular interactions of HMG14/17 with the chromatin, their function remains obscure. To address the question of the cellular function of these proteins, we have chosen.to ectopically overexpress chicken HMG14a and HMG17 in tissue culture and measure the changes, if any, in cell characteristics, including chromatin structure. Overexpression of mutant and.wild type HMG14a and HMG17 proteins may cause the exogenoquproteins to occupy new sites in the cellular chromatin and, in the case of mutants, to outcompete the endogenous HMG for binding sites. Aberrant binding of the mutant or wild type HMG proteins may then produce changes in chromatin structure. Analysis of dominant phenotypes produced in these cell lines will help to 184 establish the role these proteins play in chromatin structure and regulation of gene expression. Our studies focus on the bulk chromatin structure and make use of endonucleases that cleave at predictable locations in chromatin. Parameters such as nucleosome repeat length, spacing, sensitivity to nucleases, and changes in the structure or distribution of active and inactive sequences are determined in cell lines shown to ectopically over- express HMG14a and HMG17. 185 RESULTS Micrococcal nuclease digestion of chromatin Micrococcal nuclease is an endonuclease that digests DNA at locations not tightly complexed with proteins. In chromatin, this enzyme cuts in the linker region between nucleosomes resulting in the release of individual nucleosomes with the length of linker DNA present dependent upon the extent of digestion. DNA extracted from.a mild mdcrococcal nuclease digestion of chromatin and run out on an electrophoretic gel results in a ladder pattern with bands representing mononucleosomes, dinucleosomes, trinucleosomes and so on. The core nucleosome contains approximately 140-160 bp of DNA wrapped tighty around the octomer of core histones. The length of the linker DNA between nucleosomes is variable, usually 60-140 bp, depending upon the organism, cell type and developmental stage (75). The repeat 1ength.of:nucleosomes is the total length of the DNA associated with that nucleosome, i.e. the total of the core plus linker DNA. The spacing of nucleosomes is usually consistent within a given cell and depends on the length of linker DNA between nucleosomes. It has been suggested that HMG14/17 are required for the proper spacing of nucleosomes (82). Digestion of chromatin with micrococcal nuclease and analysis of the resultant DNA will thus demonstrate the repeat length and nucleosome spacing properties characteristic of a.particular cell line. 186 In addition, the overall sensitivity of the chromatin to digestion with micrococcal nuclease will be characteristic of a particular cell line. Regions of transcriptionally active sequences demonstrate an increased sensitivity to nucleases such as DNaseI. That the HMG14/17 proteins are also associated with these regions of active chromatin suggests they may influence this property. Figure 1 shows micrococcal nuclease digestions of chromatin from five experimental cell lines. Nuclei were isolated, digested with two concentrations of micrococcal nuclease, the DNA was extracted and then it was electrophoresed on a 1.8% agarose gel. QT6 is the untransfected control cell line, 14-RS-6 and 14-RS-8 are two single colony clones from QT6 transfected with the HMG14a cDNA containing a 3 ' -untranslated region deletion. 14-RS-6 is known to overexpress HMG14a mRNA but not protein. 178—7 and 178-10 are two single colony clones from QT6 transfected with wild type chicken HMG17 cDNA. These two clones express elevated levels of HMG17 specific mRNA as well as 3-5 times more HMG17 protein than the control cells. There is no apparent difference seen in the repeat length or spacing properties of the nucleosomes from these different cell lines. In addition, the overall nuclease sensitivity appears to be similar in the cells tested. This result is somewhat surprising, particularly in cell lines 178-7,10 which overexpress HMGl7. This 187 Figure 1 - Micrococcal nuclease digestion of chromatin DNA extracted from chromatin was digested with 50 and 300 units of micrococcal nuclease per mg nuclei DNA equivalent. The DNA. was electrophoresed. on. a 1.8% agarose gel and visualized with ethidium bromide. QT6 is the untransfected parental cell line. 14-RS-6 and 14-RS-8 are two single colony clones from a QT6 transfection of chicken HMG14a cDNA with a deletion in the 3'- untranslated region. 178-7 and 17S-10 are two single colony clones from a QT6 transfection of chicken HMG17 wild type cDNA. There is no apparent difference seen in the nucleosomal repeat length, nucleosome spacing or overall nuclease sensitivity among these clones. 188 HMG14aRS-6 HMG14 RS»8 HMG17S~7 HMG17S-10 HMG14 w4 m . mm m m.mm m 4 5 9. 4.42 Kb 4.42 Kb Figure 1 232 Figure 2 - Northern analysis of cell lines transfected with 83.38 and 83.2 cDNAs 30 ug of total RNA isolated from cell lines transfected with the H3.3B and H3.2 cDNAs in vector TFANEO was run out on 1.2% agarose/formamide gels. The gels were blotted to nitrocellulose and probed with a EcoRI LTR fragment from TFANEO. Cell lines tested include: untransfected QT6 control, QT6 transfected.with vector alone, and cell lines transfected with H3.3B and H3.2 in the sense and antisense direction in TFANEO. Bands visible in the lanes representing H3.38 and H3.2 cloned in the sense direction demonstrate that these cell lines are expressing high levels of exogenous message. 233 mBO mU> m4 u mm.mm w u mm.mm md - N.mm m . N.mm ~1300 nt ~1000 nt .8 ll. Figure 2 234 the 5' end of the exogenous histone variant mRNAs and should be unique in this cell line. Figure 2 shows a northern analysis of various cell lines probed.with the LTR fragment. Cell lines tested include the parental QT6, VC3 which is QT6 transfected with the vector alone, and H3.3B and H3.2 in the sense and antisense direction in TFANEO. The results show that the cell lines transfected with H3.3B and H3.2 in the sense orientation express reasonably high levels of exogenous message. Discussion The results presented above demonstrate that the cell lines we transfected.with histone variants H3.3B and H3.2 have the transfected cDNAs integrated into the cellular genome and express significant levels of exogenous mRNA. Microscopic examination.of these cell lines shows no gross morphological alterations and their growth rate is similar to the parental QT6. It thus appears that these cells can tolerate high levels of exogenously transfected histone H3.38 and H3.2 message with no obvious defects. Further analysis of these cell lines is required to determine if the exogenous histone mRNA is being transcribed efficiently into protein. In addition, studies of the chromatin structure in these cell lines may uncover changes brought about by the unregulated expression of these histone variants.