Higher Frequency of Concerted Evolutionary Events in Rodents Than in Man at the Polyubiquitin Gene VNTR Locus

Corresponding author: Mitsuru Nenoi, Division of Biology and Oncology, National Institute of Radiological Sciences, 9-1, Anagawa-4-chome, Inage-ku, Chiba-shi 263 Japan, m_nenoi{at}nirs.go.jp (E-mail).

Communicating editor: W.-H. LI

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS AND DISCUSSION LITERATURE CITED

The polyubiquitin gene is an evolutionarily conserved eukaryotic gene, encoding tandemly repeated multiple ubiquitins, and is considered to be subject to concerted evolution. Here, we present the nucleotide sequences of new alleles of the polyubiquitin gene UbC in humans and CHUB2 in Chinese hamster, which encode a different number of ubiquitin units from those of previously reported genes. And we analyze the concerted evolution of these genes on the basis of their orthologous relationship. That the mean of the synonymous sequence difference K_s, which is defined as the number of synonymous substitution relative to the total number of synonymous sites, within the UbC and CHUB2 genes (0.192 ± 0.096) is significantly less than K_s between these genes (0.602 ± 0.057) provides direct evidence for concerted evolution. Moreover, it also appears that concerted evolutionary events have been much more frequent in CHUB2 than in UbC, because K_s within CHUB2 (0.022 ± 0.018) is much less than that within UbC (0.362 ± 0.192). By a numerical simulation, postulating that the major mechanism of concerted evolution in polyubiquitin genes is unequal crossing over, we estimated the frequency of concerted evolutionary events of CHUB2 at 3.3 x 10^-5 per year and that of UbC at no more than 5.0 x 10^-7 per year.

UBIQUITIN is a highly conserved small protein of 76 amino acids functioning in the selective proteolysis of a variety of cellular proteins at the 26S proteasome (HOCHSTRASSER 1996 ; CIECHANOVER and SCHWARTZ 1994 ). In addition, ubiquitin has been shown to function in proteasome action-independent processes such as DNA repair (BREGMAN et al. 1996 ), endocytosis of cell surface proteins (HICKE and RIEZMAN 1996 ), and NFB signal transduction (CHEN et al. 1996 ). In humans, ubiquitin is encoded by a multiple gene family composed of UbA₅₂ , UbA₈₀ , UbB, and UbC (BAKER and BOARD 1991 , BAKER and BOARD 1987 ; WIBORG et al. 1985 ). The UbB and UbC genes, located on chromosome 17 (17p11.1-17p12) (WEBB et al. 1990 ) and chromosome 12 (12q24.3) (BOARD et al. 1992 ), respectively, are termed polyubiquitin genes because they encode tandemly repeated multiple ubiquitins with no intervening spacer. The polyubiquitin genes are conserved also in rodents, and we previously isolated and sequenced the Chinese hamster polyubiquitin genes, CHUB1 and CHUB2, which are the counterparts of the human UbB and UbC, respectively (NENOI et al. 1992 ; NENOI et al. 1994 ). Due to unequal crossing over, the number of ubiquitin units encoded by the UbC gene is variable among individuals, with the most frequent allele encoding nine units, and the other less frequent alleles encoding eight or seven units (BAKER and BOARD 1989 ). The nucleotide sequence GNNGTGGG, which has been found to be the consensus marker sequence for a variable number of tandem repeat (VNTR) loci in humans (NAKAMURA et al. 1988 ), is preserved at the 3' end of every ubiquitin-coding unit of the UbC. We have actually observed that the UbC gene alleles of HeLa cells (NENOI et al. 1996 ) as well as the Chinese hamster polyubiquitin gene CHUB2 of V79 fibroblasts (NENOI et al. 1994 ) are heterogeneous in the repeat number of the ubiquitin units.

It has been suggested that the polyubiquitin genes may show strong evidence of concerted evolution (SHARP and LI 1987 ; TAN et al. 1993 ; KEELING and DOOLITTLE 1995 ; VRANA and WHEELER 1996 ). This is consistent with observations of a high variability in the number of the ubiquitin-coding units in the UbC and CHUB2 genes because unequal crossing over is believed to be one of the major mechanisms for concerted evolution (DOVER 1982 ; DARNELL et al. 1990 ). A greater degree of homology between repeat units within a species as compared to orthologous repeat units in different species is a diagnostic of concerted evolution. However, in the case of the polyubiquitin gene, data have not been available for all loci in each species, and it has not been clear whether the loci compared across species are orthologues or paralogues.

In this report, we will analyze the unequal crossover events that are thought to have occurred on the human UbC gene and the Chinese hamster CHUB2 gene, and we will analyze the concerted evolution of these genes on the basis of their orthologous relationship. It will be proposed that the UbC gene that encodes nine ubiquitins was recently generated by an unequal crossover event between a site in the seventh ubiquitin-coding unit and the homologous site in the sixth unit of the UbC gene that encodes eight ubiquitins. We will also show a much higher homology between the ubiquitin-coding units within the CHUB2 gene than within the UbC gene, which could be explained by a higher frequency of unequal crossover events during the evolution of the CHUB2.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS AND DISCUSSION LITERATURE CITED

Isolation and sequencing of the polyubiquitin genes:
HeLa cells and V79 Chinese hamster lung fibroblasts were cultured in Eagle's MEM (Nissui, Tokyo, Japan) supplemented with 10% FBS (GIBCO BRL, Rockville, MD). Fresh thymus from a Chinese hamster was generously provided by Dr. Mitsuhiro Numata (National Institute of Health, Japan). The method for isolation of the polyubiquitin genes has been described (NENOI et al. 1996 ). Briefly, the genomic DNA was extracted from cell suspension, and the DNA fragment containing the whole ubiquitin-coding region was amplified by PCR using LA Taq polymerase (TAKARA, Otsu, Japan). Primers specific to the 5' and 3' flanking region of the UbC gene were designed on the basis of the reported nucleotide sequences (WIBORG et al. 1985 ; NENOI et al. 1996 ); P.Hu1: 5'-TTGGGCAGTGCACCCGTACCTTTGG -3', P.Hu2: 5'-GTGC AATGAAATTTGTTGAAACCTTAAAAGGGG -3', and of the CHUB2 gene (NENOI et al. 1994 ); P.CH1: 5'-CCACGAATAT TTGTCATTCCTGACCTG -3', P.CH2: 5'-GCTAAAACGAGAT CCAACACCTTTGGG -3' (indicated in Figure 2). A series of deletion mutants was constructed by partially digesting the PCR products either with PvuII (for UbC) or with Bgl II (for CHUB2), followed by subcloning in the pUC18 and then sequencing with an automated DNA sequencer (model 373A; Applied Biosystems, Foster City, CA).

View larger version (37K): In this window In a new window Download PPT slide   Figure 1. Heterogeneity in the PCR products. The polyubiquitin gene UbC of HeLa cells (A), the CHUB2 of V79 Chinese hamster cells (B), and the CHUB2 gene of the cells in Chinese hamster thymus (C) were amplified by PCR with the primer sets of P.Hu1/P.Hu2 (for UbC) and of P.CH1/P.CH2 (for CHUB2) as described under MATERIALS AND METHODS.

View larger version (65K): In this window In a new window Download PPT slide   Figure 2. Nucleotide sequence of the polyubiquitin genes. (A) the UbC of HeLa cells encoding eight ubiquitins (deposited in the DDBJ, EMBL, and GenBank nucleotide sequence databases under the accession number AB003730), (B) the CHUB2 of V79 cells encoding 13 ubiquitins (accession number AB003732) and (C) the CHUB2 of Chinese hamster thymus encoding 11 ubiquitins (accession number AB003731). Sequences are numbered from the A of the first ubiquitin initiation codon. The nucleotide sequence of the first ubiquitin-coding unit is given in full, and in the following repeats, nucleotide identity with the first unit is indicated by a dash. Primer annealing sites are underlined, and a pair of inverted repeats is boxed and designated as IR. The codon that causes amino acid substitution from Ile to Val is wavy-underlined.

Sequence analysis:
The homology analysis between every pair of ubiquitin-coding units in the UbC and CHUB2 was carried out by evaluating the synonymous sequence difference per site, K_s, which is defined as the number of synonymous substitution relative to the total number of synonymous sites (MIYATA and YASUNAGA 1980 ). K_s values were calculated as described by MIYATA and YASUNAGA 1980 and were corrected for multiple substitution (KIMURA and OHTA 1972 ).

Numerical simulation for concerted evolution of the polyubiquitin gene:
The synonymous sequence difference between several mammalian species excluding rodents has been estimated at 0.47 on average from 11 different genes (KIKUNO et al. 1985 ). Therefore, the mean evolutionary rate of synonymous substitution can be calculated to be 2.0 x 10^-9 per site per year by assuming that the mammalian divergence is 120 MYA (EASTEAL et al. 1995 ). The synonymous sequence difference between rodents and other mammals has been estimated at 0.66 on average from 35 different genes (MIYATA et al. 1987 ). Then the mean evolutionary rate of rodent genes can be calculated to be 3.5 x 10^-9 per site per year. Applying these values to polyubiquitin genes, and postulating that a major mechanism of concerted evolution is unequal crossing over (BLACK and GIBSON 1974 ; OHTA 1976 ), we constructed a Monte Carlo simulation model code in C-language on a Sun Sparc Center2000 computer. This code simulates the evolutionary change of K_s between the ubiquitin-coding units within a given polyubiquitin gene after the divergence of human and rodents. Let K_s(i,j;t) be the synonymous sequence difference between the i-th and j -th ubiquitin unit of a given polyubiquitin gene at a time t. By definition,

If a unit-duplication event occurs at the n-th unit at this moment,

On the other hand, if a unit-deletion event occurs at the n-th unit,

These calculations were repeated for the time from the man-rodent split (t = 0) until the present (t = 1.2 x 10⁸).

The sequence data presented in this article have been submitted to the DDBJ/EMBL/GenBank databases under accession numbers AB003730, AB003731, and AB003732.

	RESULTS AND DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS AND DISCUSSION LITERATURE CITED

Polymorphism of UbC and CHUB2 produced by unequal crossing over:
The UbC gene in HeLa cells and the CHUB2 gene in V79 cells and Chinese hamster thymus were amplified by PCR (Figure 1). The 2.5-kb fragments amplified from the HeLa DNA contain the previously reported UbC gene that encodes nine ubiquitins (WIBORG et al. 1985 ; EINSPANIER et al. 1987 ; NENOI et al. 1996 ). The 2.1-kb fragment amplified from the V79 DNA corresponds to the CHUB2 gene that encodes eight ubiquitins, followed by an apparently deleted and mutated ubiquitin-like polypeptide of 50-aa (NENOI et al. 1994 ). The other three products (the 2.3-kb fragment from HeLa, the 3.3-kb fragment from V79, and the 2.8-kb fragment from the Chinese hamster thymus) were isolated and sequenced. As shown in Figure 2, every product encoded tandemly repeated ubiquitins, except that an Ile was substituted with Val at the 61st amino acid position in the eighth unit of the 3.3-kb fragment from V79 cells (wavy underline in Figure 2B). Both of the 5' and 3' flanking regions precisely coincided with those of the reported UbC (WIBORG et al. 1985 ; EINSPANIER et al. 1987 ; NENOI et al. 1996 ) and CHUB2 gene sequence (NENOI et al. 1994 ), indicating that these are the allele variants of the UbC and CHUB2 in the repeat number of the ubiquitin unit. Hereafter, the polyubiquitin gene of HeLa cells encoding eight ubiquitins (Figure 1A) is designated as UbC(8u), the gene of the V79 cells encoding 13 ubiquitins (Figure 1B) is designated as CHUB2(13u), and the gene of the Chinese hamster thymus encoding 11 ubiquitins (Figure 1C) is designated as CHUB2(11u). The previously reported human UbC gene encoding nine ubiquitins and the Chinese hamster CHUB2 gene encoding eight ubiquitins are designated as UbC(9u) and CHUB2(8u), respectively.

Comparing the nucleotide sequences of the UbC gene isolated from their independent origins, BAKER and BOARD 1989 have identified an unequal crossover event site at 40–50 nt in the seventh and the ninth unit of the UbC gene encoding nine ubiquitins, that might have generated the UbC gene encoding seven ubiquitins. The sequence data they have used for comparison, however, seem very likely to have been deduced by a misalignment of the subcloned XhoI fragments (NENOI et al. 1996 ). In fact, the region of 40–50 nt, which they marked coincides with the XhoI site (BAKER and BOARD 1989 ), where the discontinuity of the sequence is thought to have been artificially introduced by the possible misalignment of the XhoI fragments. We compared the sequences of the UbC(8u)gene and the UbC(9u) gene (Figure 3), showing the position of the nucleotide discrepancy. The region from the first through the seventh unit of the UbC(9u) gene matches well with that of the UbC(8u) gene (Figure 3A). A considerable number of discrepancies can be observed from the 41 nt in the eighth unit of the UbC(9u) gene (9u-8-41) to the 3' end of the last unit. However, the region from the 220 nt in the sixth unit of the UbC(9u) gene (9u-6-220) to the 3' end of the last unit matches well with the region from the 220 nt in the fifth unit of the UbC(8u) gene (8u-5-220) to the last unit (Figure 3B). Therefore it appears that either the UbC(8u) gene would have been generated by an unequal crossover resulting in the deletion of one unit of the ubiquitin-coding sequence somewhere from (9u-5-220) to (9u-9-41) in the UbC(9u) gene (Figure 3C), or that the UbC(9u) gene was generated by the insertion of one unit of ubiquitin-coding sequence somewhere from (8u-6-220) to (8u-8-41) (Figure 3D). It is not clear which scenario is more likely. However, Table 1A shows the extremely low level of sequence difference between the seventh and both the sixth and eighth units of the UbC(9u) gene (0.059 and 0.019, respectively) compared with the differences between the other units (0.362 on average). This is what should be expected if the UbC(9u) gene was recently generated by an unequal crossover event between a site in the seventh ubiquitin-coding unit and the homologous site in the sixth unit of the UbC(8u) gene (case in Figure 3D).

View larger version (54K): In this window In a new window Download PPT slide   Figure 3. Comparison of the nucleotide sequences between the UbC(8u) and the UbC(9u), and plausible mechanisms for unequal crossing over. Ubiquitin-coding units are aligned one by one from the 5' end (A) and from the 3' end (B). Sites of nucleotide discrepancies are marked by circles. (C) Illustration showing a possible mechanism for the generation of the UbC(8u) allele by an unequal crossing over between a pair of UbC(9u) alleles. The possible event is considered to have occurred at a site in the hatched region. (D) Illustration showing a possible mechanism for the generation of the UbC(9u) allele by an unequal crossing over between a pair of UbC(8u) alleles.

Additionally, a very low sequence difference between the second and the fourth units of the UbC(9u) gene was observed (Table 1A). Relatively low sequence differences between the third and both the first and fifth units were also observed. This also is to be expected if two ubiquitin-coding units were inserted by an unequal crossing over between a site in the third unit and the homologous site in the first unit of the ancestral UbC gene. This suggests that such an event must have occurred before the UbC(9u) gene was ever generated.

In contrast to the human UbC gene, it seemed nearly impossible to estimate the site of unequal crossover events in the CHUB2 gene only by comparing the nucleotide sequences of the CHUB2(8u), CHUB2(11u), and CHUB2(13u) because of the extremely low level of sequence differences among the ubiquitin-coding units in each of these alleles (Table 1B, data not shown). The small sequence differences within the CHUB2 alleles seem to be caused by concerted evolution.

Evidence for the concerted evolution of polyubiquitin genes:
As multiple polyubiquitin gene loci have been observed in mammals, distinguishing between orthologous and paralogous homology is of great importance when analyzing concerted evolution by comparing the degree of homology between repeats both in different species and within a locus (SHARP and LI 1987 ; VRANA and WHEELER 1996 ). However, in previous studies dealing with the concerted evolution of the polyubiquitin gene in mammals, only the presence of a higher level of homology within a species than between species has been used as evidence of concerted evolution (SHARP and LI 1987 ; TAN et al. 1993 ; KEELING and DOOLITTLE 1995 ).

We have previously identified a high degree of homology in the 3' UTR between the UbC and the CHUB2 (74 matches out of 88 bp when gaps were introduced), and showed also that they share a pair of inverted repeats of 10 bp in length at the same location with the same sequence (NENOI et al. 1994 ) (boxes designated as IR in Figure 2). In addition, we isolated another Chinese hamster polyubiquitin gene, CHUB1, which encodes five units, and showed both that its 3' UTR is highly homologous with that of the human polyubiquitin gene UbB (147 matches out of 165 bp when gaps were introduced) and that a pair of inverted repeats is conserved at a different location and with a different nucleotide sequence from those of the UbC and the CHUB2 (NENOI et al. 1992 ). Based on these facts, it is quite reasonable to assume that both the relationship between the UbC and the CHUB2 and between the UbB and the CHUB1 are orthologous. This assumption is further supported by the data in Table 2, which shows that the sequence differences between UbC and CHUB2 and those between UbB and CHUB1 are evidently smaller than those between UbC and CHUB1 as well as those between UbB and CHUB2.

Consequently, the evidence for concerted evolution can be directly shown by indicating that the mean of the sequence difference within the UbC(9u) gene and that within the CHUB2(11u) gene {[(0.362 ± 0.192) + (0.022 ± 0.018)]/2 = 0.192 ± 0.096 per site, Table 1, A and B} is significantly less than the sequence difference between these genes (0.602 ± 0.057 per site, Table 2). We used the alleles of the UbC(9u) and CHUB2(11u) for the present analysis because these alleles are present in individuals, and are thought to be least affected by artificial procedures such as cell culturing. This is the first direct evidence for concerted evolution deduced from a comparison between a pair of orthologous polyubiquitin genes in mammals. In addition, it is evident that concerted evolutionary events have been much more frequent in the CHUB2 gene than in the UbC gene because the sequence difference within the CHUB2(11u) gene (0.022 ± 0.018 per site) is much smaller than that within the UbC(9u) gene (0.362 ± 0.192 per site), in spite of a higher rate of synonymous substitutions in rodents than in man (WU and LI 1985 ; KIKUNO et al. 1985 ).

The sequence differences within the human UbB and the Chinese hamster CHUB1 were also estimated (Table 3). Again, evidence of concerted evolution was apparent {[(0.187 ± 0.100) + (0.067 ± 0.035)]/2 = (0.127 ± 0.053) < (0.472 ± 0.061)}, and it appeared that there was a higher frequency of concerted evolutionary events in the CHUB1 gene than in the UbB gene [(0.067 ± 0.035) < (0.187 ± 0.100)]. A similarly low level of sequence difference is observed (K_s = 0.035 ± 0.042 per site) also within a rat polyubiquitin gene (HAYASHI et al. 1994 ), suggesting this is a common feature in rodents.

There are three possible reasons for such high sequence similarity within the Chinese hamster polyubiquitin genes, and possibly in other rodents as well. First, there may be a strong evolutionary constraint on the nucleotide sequence of the polyubiquitin genes in rodents, and the polyubiquitin gene alleles harboring a nucleotide substitution may have been removed from the population. However it is unlikely that such constraints are imposed only on the genes in rodents because the polyubiquitin gene is commonly conserved in all eukaryotes. In fact, the synonymous sequence differences between the UbC gene and the CHUB2 gene (0.602 ± 0.057 per site) demonstrate that both genes have evolved with a comparable rate of synonymous substitution to that of other genes. Second, the concerted evolutionary events may have occurred very recently in the polyubiquitin genes in rodents. However the observed sequence similarity involving all units of the CHUB2 has obviously required a series of events. It is very unlikely that these multiple events occurred together at once. Third, the frequency of the concerted evolutionary event in the polyubiquitin genes may have been higher especially in rodents than in other mammals. We tested this possibility for the CHUB2 gene with the numerical simulation shown below.

Numerical simulation of the concerted evolution:
Generally, concerted evolution in tandemly repeated genes is believed to occur by means of one of two mechanisms: frequent unequal crossing over or gene conversion (DARNELL et al. 1990 ). SHARP and LI 1987 have suggested the possible involvement of gene conversion to explain a high sequence similarity between the second and the eighth ubiquitin-coding unit in the previously reported UbC gene sequence (WIBORG et al. 1985 ). However, again, their analysis was based on sequence data derived from a probable misalignment of the XhoI fragments. In the UbC gene sequence that we determined, these units are located closely to each other at the second and the fourth units (NENOI et al. 1996 ). The difference between species in the repeat number of ubiquitin-coding units per locus (NENOI et al. 1994 ; NENOI et al. 1992 ), and evidence of within-population polymorphism in repeat number (BAKER and BOARD 1989 ; NENOI et al. 1996 ), point to a frequent occurrence of unequal crossing over within these loci.

By analogy to the analyses of concerted evolution in multigene families carried out by BLACK and GIBSON 1974 and OHTA 1976 , it can be considered that the repeated duplication and deletion of the ubiquitin-coding units by unequal crossing over may have eventually resulted in the high sequence homology observed in the present CHUB2 gene. We have indicated this possibility by numerically simulating K_s, assuming that unequal crossover events have occurred randomly among all units and randomly in time, and that the duplication and deletion have occurred by one ubiquitin-coding unit, as supporsed by OHTA 1976 . We used K_s for the present analysis because K_s is directly related to the evolutionary distance between the species (or genes, loci) being compared, irrespective of gene type and location on the chromosome (MIYATA et al. 1980 ). Therefore, it is possible to apply the equal rate of synonymous substitution to that derived from a variety of rodent genes (3.5 x 10^-9 per site per year) to the case of CHUB2 gene evolution. This can be rationalized by the observation that the sequence difference between the CHUB2 and the UbC [0.602 ± 0.057 per site (Table 2)] is very close to that estimated for other genes between rodents and human (0.66 ± 0.13 per site). This close correlation implies that most of the regions over the ubiquitin-coding unit of the UbC and the CHUB2 genes have not been under any strong evolutionary constraint since the divergence of man and rodents, and that synonymous substitutions have been accumulated at a rate common to that of other genes (2.0 x 10^-9 per site per year for genes of mammals other than rodents, 3.5 x 10^-9 per site per year for rodent genes). This is further supported by the observation of MITA et al. 1991 that the codon choices among synonymous codons in polyubiquitin genes generally follows the G+C content of the overall coding regions in corresponding organisms.

Our simulation code can provide an estimate of the sequence difference within a polyubiquitin gene after an arbitrary number of unequal crossover events since the divergence of man and rodents [K_s(i,j;t = 1.2 x 10⁸)], provided that the repeat number and the sequence differences within the ancestral polyubiquitin gene are given as the initial condition. We, however, have no available data for such an initial condition. The simulation was then carried out with the initial unit number, N₀ = 1, 10, and 20, and the initial sequence differences, K_s(i,j; 0) = 0 for all i and j (1 i, j N₀). Adoption of a zero initial sequence difference implies that this simulation should give the minimum estimation for K_s(i,j;t = 1.2 x 10⁸), and therefore the minimum estimation for the frequency of unequal crossover events. Independent simulations were carried out 100 times for all of the event numbers and initial conditions. Figure 4A shows the sequence differences at t = 1.2 x 10⁸, averaged for all unit combinations (designated as K_s(t = 1.2 x 10⁸) in the figure). It is evident that K_s(t = 1.2 x 10⁸) decreases with an increase in the event number N_event , and that K_s(t = 1.2 x 10⁸) is not affected by the initial unit number N₀ when N_event 200. These results suggest that, to explain the extremely low sequence differences within the present CHUB2(11u) gene (0.022 ± 0.018, the horizontal line in Figure 4A), unequal crossover events must have occurred at least 4000 times in the CHUB2(11u) gene since the divergence of human and rodents 120 MYA (3.3 x 10^-5 per year). Figure 4B shows the dependency of K_s(t = 1.2 x 10⁸) on the initial sequence difference K_s(i,j;t = 0) (a common number was postulated for all i and j (1 i, j N₀), designated as K_s(t = 0) in the figure), when N₀ and N_event were fixed to 10 and 3999, respectively. It can be seen that K_s(t = 1.2 x 10⁸) is no longer affected by K_s(t = 0) after such frequent unequal crossing over as N_event = 3999. Therefore it is reasonable to conclude that the deduced value of 3.3 x 10^-5 per year is not the minimum estimate, but is now the expected value for the frequency of unequal crossover event that occurred on the CHUB2(11u) gene.

View larger version (17K): In this window In a new window Download PPT slide   Figure 4. Simulation for the sequence difference within the CHUB2(11u) gene. (A) The sequence difference Ks(t = 1.2 x 108) after various number of unequal crossover events Nevent is plotted, when the rate of synonymous substitution v is assumed to be 3.5 x 10-9 per site per year, and the sequence difference within the ancestral polyubiquitin gene Ks(t = 0) is postulated to be 0. The unit number of the ancestral polyubiquitin gene N0 is postulated as in the box. The level of the sequence difference observed for the present CHUB2(11u) gene is indicated by a horizontal line. (B) The sequence difference after 3999 unequal crossover events is plotted, when various sequence differences are postulated for the ancestral polyubiquitin gene. Error bars represent the standard deviation.

The simulation was also carried out for the UbC(9u) gene, postulating v = 2.0 x 10^-9 with the same initial conditions as those used for the CHUB2 gene [N₀ = 1, 10, 20, K_s(t = 0) = 0]. Figure 5A shows the averaged sequence difference deduced from 100 independent simulations. In this case, K_s(t = 1.2 x 10⁸) comes close to the observed sequence difference for the present UbC(9u) gene (0.362 ± 0.192 per site) with an event number N_event of no more than 60 (5.0 x 10^-7 per year) even when the initial unit number N₀ was assumed to be 20. The simulation of K_s(t = 1.2 x 10⁸) does not significantly depend on K_s(t = 0) if K_s(t = 0) is less than one (Figure 5B), which is a reasonable assumption considering that polyubiquitin genes have been subject to concerted evolution even before mammalian divergence.

View larger version (25K): In this window In a new window Download PPT slide   Figure 5. Simulation for sequence differences within the UbC(9u) gene. (A) The sequence difference Ks(t = 1.2 x 108) after various numbers of unequal crossover events Nevent is plotted, when the rate of synonymous substitution v is assumed to be 2.0 x 10-9 per site per year. The level of sequence difference observed for the present UbC(9u) gene is indicated by a horizontal line. (B) The dependence of Ks(t = 1.2 x 108) on Ks(t = 0) for various numbers of unequal crossover events. Error bars represent the 0.2x standard deviation.

The frequency of unequal crossover events was also numerically estimated for the human UbB gene and the Chinese hamster CHUB1 gene (Figure 6). In this case, the simulation was carried out under the constraint that the unit number is kept between one and eight. Again, it is evident that the frequency of unequal crossing over was higher in the Chinese hamster CHUB1 (1.7 x 10^-6 per year) than in the human UbB (3.3 x 10^-7 per year). Lower frequencies of unequal crossing over in CHUB1 and UbB than those in CHUB2 and UbC are considered to be due to a smaller number of ubiquitin-coding units in these genes.

View larger version (21K): In this window In a new window Download PPT slide   Figure 6. Simulation for sequence differences within the Chinese hamster CHUB1 gene (A) and the human UbB gene (B). The sequence difference Ks(t = 1.2 x 108) after various numbers of unequal crossover events Nevent is plotted, when the rate of synonymous substitution v is assumed to be 3.5 x 10-9 per site per year for CHUB1 and 2.0 x 10-9 per site per year for UbB. The level of sequence difference observed for the present polyubiquitin gene is indicated by a horizontal line. Error bars represent the standard deviation.

Concerted evolution has been thoroughly investigated for the family of rRNA genes. In humans, the 45S pre-rRNA gene and 5S rRNA gene are present in ~250 and ~2000 copies, respectively, in tandem arrays (DARNELL et al. 1990 ). STRACHAN et al. 1985 have estimated the frequency of unequal crossover events in the rRNA genes of Drosophila at 10^-2–10^-4 per generation. Compared with this value, together with the consideration that our present simulation may have overestimated the frequency of unequal crossing over in the CHUB2(11u) gene because of neglecting the involvement of gene conversion, it can be noted that the unequal crossover events have occurred much less frequently in the CHUB2(11u) gene than in the rRNA genes. This agrees with the observation of SHARP and LI 1987 that the rate of concerted evolution seems to be higher in the rRNA gene family than in the ubiquitin genes of various eukaryotes although the organization of the rRNA gene array is considered to be less conducive to concerted evolution than that of a polyubiquitin locus.

A simple explanation for the different event-frequency between man and rodents is that there would be a generation time effect on the frequency of unequal crossover events. From the study of the human minisatellite loci, it has been pointed out by JEFFREY et al. (1988) that the large mutation events, involving the gain or loss of up to about 2 kb, appear to arise by means of recombinational processes at meiosis. As the generation time of rodents is considered to be 100 times shorter than that of human, it is reasonable to speculate that rodents have been more susceptible to unequal crossing over during meiosis since the divergence of these species.

	ACKNOWLEDGMENTS

The authors thank Dr. NARUYA SAITOU of National Institute of Genetics of Japan, for his helpful suggestions and comments on this study.

Manuscript received July 28, 1997; Accepted for publication October 24, 1997.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS AND DISCUSSION LITERATURE CITED

BAKER, R. T. and P. G. BOARD, 1987  The human ubiquitin gene family: structure of a gene and pseudogenes from the Ub B subfamily. Nucleic Acids Res. 15:443-463[Abstract/Free Full Text].

BAKER, R. T. and P. G. BOARD, 1989  Unequal crossover generates variation in ubiquitin coding unit number at the human UbC polyubiquitin locus. Am. J. Hum. Genet. 44:534-542[Medline].

BAKER, R. T. and P. G. BOARD, 1991  The human ubiquitin-52 amino acid fusion protein gene shares several structural features with mammalian ribosomal protein genes. Nucleic Acids Res. 19:1035-1040[Abstract/Free Full Text].

BLACK, J. A. and D. GIBSON, 1974  Neutral evolution and immunoglobulin diversity. Nature 250:327-328[Medline].

BOARD, P. G., M. COGGAN, R. T. BAKER, J. VUUST, and G. C. WEBB, 1992  Localization of the human UBC polyubiquitin gene to chromosome band 12q24.3. Genomics 12:639-642[Medline].

BREGMAN, D. B., R. HALABAN, A. J. VAN GOOL, K. A. HENNING, and E. C. FRIEDBERG et al., 1996  UV-induced ubiquitination of RNA polymerase II: a novel modification deficient in Cockayne syndrome cells. Proc. Natl. Acad. Sci. USA 93:11586-11590[Abstract/Free Full Text].

CHEN, Z. J., L. PARENT, and T. MANIATIS, 1996  Site-specific phosphorylation of IB by a novel ubiquitination-dependent protein kinase activity. Cell 84:853-862[Medline].

CIECHANOVER, A. and A. L. SCHWARTZ, 1994  The ubiquitin-mediated proteolytic pathway: mechanisms of recognition of the proteolytic substrate and involvement in the degradation of native cellular proteins. FASEB J. 8:182-191[Abstract].

DARNELL, J., H. LODISH and D. BALTIMORE, 1990 Eukaryotic chromosomes and genes: molecular anatomy, pp. 347–388 in Molecular Cell Biology. W. H. Freeman and Company, New York.

DOVER, G., 1982  Molecular drive: a cohesive mode of species evolution. Nature 299:111-117[Medline].

EASTEAL, S., C. COLLET and D. BETTY, 1995 The Mammalian Molecular Clock. Springer-Verlag, Heidelberg.

EINSPANIER, R., H. S. SHARMA, and K. H. SCHEIT, 1987  Cloning and sequence analysis of a cDNA encoding poly-ubiquitin in human ovarian granulosa cells. Biochem. Biophys. Res. Commun. 147:581-587[Medline].

HAYASHI, T., M. NOGA, and M. MATSUDA, 1994  Nucleotide sequence and expression of the rat polyubiquitin mRNA. Biochim. Biophys. Acta 1218:232-234[Medline].

HICKE, L. and H. RIEZMAN, 1996  Ubiquitination of a yeast plasma membrane receptor signals its ligand-stimulated endocytosis. Cell 84:277-287[Medline].

HOCHSTRASSER, M., 1996  Ubiquitin-dependent protein degradation. Annu. Rev. Genet. 30:405-439[Medline].

JEFFREYS, A. J., N. J. ROYLE, V. WILSON, and Z. WONG, 1988  Sponta-neous mutation rates to new length alleles at tandem-repetitive hypervariable loci in human DNA. Nature 332:278-281[Medline].

KEELING, P. J. and W. F. DOOLITTLE, 1995  Concerted evolution in protists: recent homogenization of a polyubiquitin gene in Trichomonas vaginalis.. J. Mol. Evol. 41:556-562[Medline].

KIKUNO, R., H. HAYASHIDA, and T. MIYATA, 1985  Rapid rate of rodent evolution. Proc. Jpn. Acad. 61(B):153-156.

KIMURA, M. and T. OHTA, 1972  On the stochastic model for estimation of mutational distance between homologous proteins. J. Mol. Evol. 2:87-90[Medline].

MITA, K., S. ICHIMURA, and M. NENOI, 1991  Essential factors determining codon usage in ubiquitin genes. J. Mol. Evol. 33:216-225[Medline].

MIYATA, T., H. HAYASHIDA, K. KUMA, and T. YASUNAGA, 1987  Male-driven molecular evolution demonstrated by different rates of silent substitutions between autosome- and sex chromosome-linked genes. Proc. Jpn. Acad. 63(B):327-331.

MIYATA, T. and T. YASUNAGA, 1980  Molecular evolution of mRNA: a method for estimating evolutionary rates of synonymous and amino acid substitutions from homologous nucleotide sequences and its application. J. Mol. Evol. 16:23-36[Medline].

MIYATA, T., T. YASUNAGA, and T. NISHIDA, 1980  Nucleotide sequence divergence and functional constraint in mRNA evolution. Proc. Natl. Acad. Sci. USA 77:7328-7332[Abstract/Free Full Text].

NAKAMURA, Y., M. CARLSON, K. KRAPCHO, M. KANAMORI, and R. WHITE, 1988  New approach for isolation of VNTR markers. Am. J. Hum. Genet. 43:854-859[Medline].

NENOI, M., K. MITA, and S. ICHIMURA, 1992  Evolutionarily conserved structure of the 3' non-translated region of a Chinese hamster polyubiquitin gene. Biochim. Biophys. Acta 1130:247-252[Medline].

NENOI, M., K. MITA, S. ICHIMURA, and I. L. CARTWRIGHT, 1994  Novel structure of a Chinese hamster polyubiquitin gene. Biochim. Biophys. Acta 1204:271-278[Medline].

NENOI, M., K. MITA, S. ICHIMURA, I. L. CARTWRIGHT, and E. TAKAHASHI et al., 1996  Heterogeneous structure of the polyubiquitin gene UbC of HeLa S3 cells. Gene 175:179-185[Medline].

OHTA, T., 1976  Simple model for treating evolution of multigene families. Nature 263:74-76[Medline].

SHARP, P. M. and W. H. LI, 1987  Ubiquitin genes as a paradigm of concerted evolution of tandem repeats. J. Mol. Evol. 25:58-64[Medline].

STRACHAN, T., D. WEBB, and G. A. DOVER, 1985  Transition stages of molecular drive in multiple-copy DNA families in Drosophila.. EMBO J. 4:1701-1708[Medline].

TAN, Y., S. T. BISHOFF, and M. A. RILEY, 1993  Ubiquitins revisited: further examples of within- and between-locus concerted evolution. Mol. Phylogenet. Evol. 2:351-360[Medline].

VRANA, P. B. and W. C. WHEELER, 1996  Molecular evolution and phylogenetic utility of the polyubiquitin locus in mammals and higher vertebrates. Mol. Phylogenet. Evol. 6:259-269[Medline].

WEBB, G. C., R. T. BAKER, K. FAGAN, and P. G. BOARD, 1990  Localiza-tion of the human UbB polyubiquitin gene to chromosome band 17p11.1-17p12. Am. J. Hum. Genet. 46:308-315[Medline].

WIBORG, O., M. S. PEDERSEN, A. WIND, L. E. BERGLUND, and K. A. MARCKER et al., 1985  The human ubiquitin multigene family: some genes contain multiple directly repeated ubiquitin coding sequences. EMBO J. 4:755-759[Medline].

WU, C. I. and W. H. LI, 1985  Evidence for higher rates of nucleotide substitution in rodents than in man. Proc. Natl. Acad. Sci. USA 82:1741-1745[Abstract/Free Full Text].

This article has been cited by other articles:

				Z. Zhang, T. Cavalier-Smith, and B. R. Green Evolution of Dinoflagellate Unigenic Minicircles and the Partially Concerted Divergence of Their Putative Replicon Origins Mol. Biol. Evol., April 1, 2002; 19(4): 489 - 500. [Abstract] [Full Text] [PDF]

				M. Nei, I. B. Rogozin, and H. Piontkivska Purifying selection and birth-and-death evolution in the ubiquitin gene family PNAS, September 26, 2000; 97(20): 10866 - 10871. [Abstract] [Full Text] [PDF]