Recombination between co-infecting herpesviruses occurs where replication compartments coalesce

Recombination between co-infecting herpesvirus DNA genomes is an important process affecting viral population diversity and evolution. Each herpes simplex virus type 1 (HSV-1) replication compartment (RC) derives from a single incoming genome and maintains a specific territory within the nucleus. This raises intriguing questions about where and when co-infecting viral genomes interact. To study the spatiotemporal requirements for inter-genomic recombination, we developed an assay with dual-color fluorescence in situ hybridization enabling detection of homologous recombination (HR) between different co-infecting HSV-1 pairs. Our results revealed that when viral RCs enlarge towards each other, there is detectable overlap of genomes from each virus. Infection with paired viruses that allow visualization of HR correlates with increased overlap of RCs. Taken together, these findings suggest that HR events take place during replication of HSV-1 DNA and are mainly confined to the periphery of RCs when they coalesce. Importance Homologous recombination is considered a major driving force of evolution since it generates and spreads genetic diversity. In the case of HSV-1, a ubiquitous human pathogen that causes significant morbidity and mortality, evidence of homologous recombination can be found frequently, both in vitro and in clinical isolates. Here we designed an experimental system to detect where and when recombination takes place between viral genomes during infection. We found that recombination events occur after viral DNA replication during the late stages of infection, and is prevalent at the interface of expanding viral replication compartments. Overall, our results provide spatial and temporal information regarding the process of HSV-1 replication and recombination, and these observations have implications for understanding the recombination restrictions of other DNA viruses and cellular DNA.


23
Recombination between co-infecting herpesvirus DNA genomes is an important 24 process affecting viral population diversity and evolution. Each herpes simplex virus 25 type 1 (HSV-1) replication compartment (RC) derives from a single incoming genome 26 and maintains a specific territory within the nucleus. This raises intriguing questions 27 about where and when co-infecting viral genomes interact. To study the 28 spatiotemporal requirements for inter-genomic recombination, we developed an assay 29 with dual-color fluorescence in situ hybridization enabling detection of homologous 30 recombination (HR) between different co-infecting HSV-1 pairs. Our results revealed 31 that when viral RCs enlarge towards each other, there is detectable overlap of 32 genomes from each virus. Infection with paired viruses that allow visualization of HR 33 correlates with increased overlap of RCs. Taken together, these findings suggest that 34 HR events take place during replication of HSV-1 DNA and are mainly confined to the 35 periphery of RCs when they coalesce.  (17). A recent study showed that viral genomes entering the nucleus are 71 observed as condensed foci, and suggested that viral expression and DNA replication 72 allow decondensation of these genomes and formation of RCs (19). Interestingly, 73 some genomes remain highly condensed at the edge of newly developing RCs (19). 74 Here, we visualised co-infecting HSV-1 genomes and confirmed that 75 alphaherpesviruses RCs initiate from single genomes. 76 Recombination is considered to be a major driving force in evolution of most organisms, 77 since it accelerates adaptation (20,21). Viruses with double strand DNA genomes need to adapt to the changing environment. Since the rate of mutation accumulation 79 is lower for DNA viruses than that of viruses with RNA genomes (22,23), it has been 80 hypothesised that high rates of recombination can facilitate genetic adaptation (24). 81 Indeed, homologous recombination (HR) among co-infecting HSV genomes is very 82 frequently observed in both in vitro genetic assays (25)(26)(27)(28)(29)(30)(31) and in sequence analysis 83 of clinical isolates (32)(33)(34). 84 Viral DNA recombination can be impacted by both viral and cellular proteins. Two viral 85 proteins have been suggested to work as a complex to facilitate viral recombination 86 and have been shown to catalyze strand exchange in vitro: the single strand binding 87 protein ICP8 and an exonuclease UL12 (35). Single strand annealing was found to be 88 a recombination mechanism upregulated during viral infection and thus is considered 89 as the mechanism by which the viral recombinase induces recombination (36). While 90 ICP8 is required for viral DNA replication, UL12 is not essential for DNA replication per 91 se, although it is required for formation of infectious viral genomes that can be 92 packaged into capsids (37). Recombination of HSV-1 genomes can be enhanced by 93 DNA double stand breaks (DSB) (31, 38) and components of cellular DSB repair 94 pathways are recruited to sites of viral DNA replication (4,6,8,39,40). Host proteins 95 that are known to be involved in host HR, were found to support either viral DNA 96 replication or production of infective viral progeny (4,39,(41)(42)(43). These findings 97 suggest that viral recombination and replication are processes that are closely related 98 (44). While knowledge regarding the molecular aspects of HR has been accumulating 99 over the last few years, little is known regarding spatiotemporal constraints on inter-100 genomic recombination.

101
The compartmentalization of co-infecting genomes at different RCs raises the question 102 of where and when recombination takes place during the course of infection. To tackle this question, we utilized a fluorescence in situ hybridization (FISH) based assay to 104 differentiate between de-novo synthesized variants of viral genomes. Our results 105 suggest that multiple recombination events occur at later stages of infection following 106 DNA replication and that recombination takes place at the interface between mature 107 RCs. We also found strong evidence of correlation between the number of RCs and 108 nuclear size, suggesting a possible spatial restriction on the number of viral genomes 109 that initiate replication.

111
A FISH based assay designed to detect recombination events between co-112 infecting viruses. Recombination among co-infecting HSV-1 strains is a frequent 113 event that can be detected by the progeny viruses released from infected cells (27)(28)(29)114 31). To study the spatiotemporal constraints of these recombination events we 115 developed a FISH based experimental assay that enables visualization of two different 116 viral genomes within a co-infected nucleus. For this assay, we constructed a series of 117 viral isolates, isogenic to each other except for two unique tag sequences (YPET or 118 mCherry, yellow and red fluorescent proteins encoding genes, respectively) inserted 119 into various loci throughout the viral genome ( Figure 1A and Table 1). We designed 120 two sets of fluorescent probes, one set for each tag sequence ( Figure 1B). Each probe 121 set was conjugated to a distinct fluorophore (Cy3 or Cy5) to enable visual identification 122 of the genomes containing each tag sequence. We hypothesize that using different 123 mixtures of the viruses should lead to distinct patterns within the infected nucleus 124 ( Figure 1C). As was shown previously for PRV (18), we expect that co-infection with 125 two HSV-1 viruses containing tag sequences at the same genomic locus cannot result 126 in a new recombinant genome containing both tags. Thus, each RC will react to a 127 single fluorophore, including at the contacting edges of proximate RCs ( Figure 1C example I). Infection with a viral recombinant containing two tag sequences within one 129 genome is expected to result in RCs stained with both fluorophores (dually-labelled 130 RCs) ( Figure 1C example IV). Infection with two viral recombinants containing tag 131 sequences at different genomic loci could result in progeny genomes that either 132 contain both tag sequences on a single genome or contain neither tag sequence. One 133 of two types of spatial patterns could be expected to dominate under these conditions.

134
Dually-labelled RCs (fully covered by both probe sets) imply that recombination takes 135 place early during infection before the viral DNA replicates and RCs mature. In this 136 situation, the reciprocal recombinant genome would contain no tag sequences and will 137 generate RCs that are not detected since they are not covered by any of the probes 138 ( Figure 1C example II). Alternatively, partially overlapping RCs that fuse to each other 139 at their periphery, suggest that recombination occurs later during the infection cycle 140 following viral DNA replication ( Figure 1C example III). Therefore, this FISH assay is  Figure 1D), suggesting that progeny viruses can be the outcome of HR events (27).

154
To visualize HR events at the single cell level, cells were co-infected and fixed at 6HPI 155 for hybridization with the appropriate fluorescent probes. Using confocal fluorescent 156 microscopy, we imaged the viral RCs and identified four distinct patterns of 157 interactions between the RCs (Figure 2A-D). The first interaction type is of two RCs 158 that come into close contact but without evidence of mixing between the two RCs i.e. 159 no visible co-localized pixels and no intersection between RCs margins (Figure 2A).

160
These interactions were mostly observed when the tag sequences were at the same 161 genomic locus (OK25 and OK35, Figure 2E). The second interaction type is of two 162 RCs that come into close contact and clearly mix, as defined by the presence of dual 163 color pixels at the site of contact or blurring of the contacting edges ( Figure 2B). This 164 was the most common interaction observed in all co-infections, particularly when the 165 two tag sequences were located further from each other on the viral genome (OK26 166 and OK35, Figure 2E). In the third interaction type, one RC seems to be fully 167 overlapped with part of a second RC that is usually much larger ( Figure 2C). This both tag sequences (OK31), and was rarely detected during co-infections. We deduce 174 that our assay can be used to readout recombination, since detectable HR events 175 change the overall distribution of the interactions among the RCs. In all co-infection 176 assays, partial overlap at the point of contact are more frequent, compared to entirely 177 overlapping RCs. The rare occurrence of entirely overlapping RCs compared to the 178 expected high frequency of HR during HSV-1 co-infection (31) suggest that these RCs 179 are unlikely to reflect most of the HR events. We therefore suggest that inter-genomic 180 HR occur at points of physical interaction between co-infecting genomes, after viral 181 DNA replication has initiated. 182 We have previously shown that in U2OS cells, on average more HSV-1 genomes HR had much lower effect, probably due to higher probability for no interactions 218 between the RCs. These results suggest that when HR is possible to detect between 219 co-infecting viruses, interactions between RCs are observed more frequently. 220 We analysed all the cells in which co-localization was observed. For both cell types 221 we found a significant increase in the relative overlapping area for each nucleus for 222 cells infected by viruses with tags in different loci ( Figure 4C-D). The resolution of our 223 FISH assay was not sufficient to detect significant differences between the two co-224 infections with viruses that have tags in different loci. To test that cellular parameters 225 did not bias our measurements, we compared the relative overlapping area to the 226 nuclear area ( Figure 4E-F) or to the number of RCs per nucleus (not shown). We found 227 no evidence of correlation between these two parameters to the relative overlapping 228 area in all infection conditions (Pearson correlation bellow 0.3 for each of the 229 infections). We conclude that the observed significant increase in overlapping area 230 indicates the readout of HR events. We speculate that the relative high background 231 levels of overlapping areas (~20%) are due in part to noise of the assay and to non-232 HR events during viral replication.  We speculated that the increase in total RCs area could also result from the possibility 251 that RCs may expand faster to a larger size in larger nuclei. We therefore compared 252 the mean RC area (per cell) to the nuclear area ( Figure 5G-H). We found much weaker 253 correlation between these parameters (Pearson correlation: < 0.5 for U20S cells and 254 < 0.3 for Vero cells in each of the infections). These results suggest that the increase 255 in RC area in larger nuclei results from higher numbers of RCs rather than increased 256 RC size. Taken together, our results suggest that nuclear size is a limiting factor for 257 the number of incoming genomes that are able to initiate replication. To determine whether these RCs exist, we designed an additional 263 probe conjugated to a third fluorophore corresponding to the genomic HSV a' 264 sequence, a repetitive sequence found in four copies within the HSV-1 genome (47), 265 This probe stains all HSV-1 viral DNA regardless of the tag sequence inserted. Both 266 Vero and U2OS cells were co-infected with two isolates containing tags at different 267 genomic loci, fixed and hybridized to all three probes as previously described. We 268 inspected over 600 RCs from each cell type and found that 99.6% stained by the HSV 269 non-specific probes reacted to either of the two specific probe sets (example in Figure   270 6). The existence of at least one tag sequence for all RCs suggests that RCs are not in agreement with the hypothesis that recombination events occur between replicating 300 genomes in either concatemeric or circular state (26). 301 We observed four recognizable patterns of interactions between adjacent RCs: i. no 302 mixing, ii. partial mixing, iii. one within another and iv. complete overlap (Figure 2 and   303 3). From our previous results with PRV (18), we expected that the no mixing of RCs 304 would be the dominant interaction following co-infection with two viruses carrying the 305 tag sequences at the same location in the viral genomes (OK25 and OK35). We found 306 that this readout is most commonly observed in this co-infection, although to our 307 surprise the majority of RCs (~66%) showed some degree of mixing. Comparison with 308 the images obtained in the Kobiler et al. paper suggested that there is no major 309 difference between the images from the two alphaherpesviruses co-infection, although 310 robust quantitative analysis of PRV images was not carried out.

311
The partial mixing of RCs was the most common interaction observed among the 312 different co-infection conditions tested ( Figure 2E). This is in part due to the into the same genome, or that there are high levels of non-HR that take place during 325 viral replication. We note that non-HR events were observed previously both in HSV-1 replication (49) and in other herpesviruses (50). These explanations are not mutually 327 exclusive, and probably both contribute to the mixing of colors observed without HR.

328
The interaction in which one RC is fully mixed into a larger RC ( Figure 2C)  RCs initiate from single incoming genomes (18). 346 We previously demonstrated that the number of HSV-1 genomes replicating per cell    The viral recombinants were validated by PCR. Viral titters were measured by plaque 392 assay. An additional viral recombinant containing both tag sequences was isolated by 393 crossing the recombinant OK26 to the previously described OK11 (57) and selecting 394 for plaques containing two fluorescent proteins by plaque assay. All viral recombinants 395 constructed for this paper are described in Table 1 396 Fluorescent probes 397 A set of 20 fluorescent probes was designed to correspond to each one of the two tag  Table 2.

441
To exclude pixels with background intensities from further analysis, the threshold was 442 applied for each nucleus mask. Otsu thresholding was then applied to the remaining 443 pixels, pixels below the threshold were pooled and a new threshold of two standard 444 deviations above the mean was calculated. The pixels above these thresholds were 445 defined as the RCs. The segmentation refined by applying morphological operators: