Alcelaphine herpesvirus 1 genes A7 and A8 regulate viral spread and are essential for malignant catarrhal fever

Alcelaphine herpesvirus 1 (AlHV-1) is a gammaherpesvirus that is carried asymptomatically by wildebeest. Upon cross-species transmission to other ruminants, including domestic cattle, AlHV-1 induces malignant catarrhal fever (MCF), which is a fatal lymphoproliferative disease resulting from proliferation and uncontrolled activation of latently infected CD8+ T cells. Two laboratory strains of AlHV-1 are used commonly in research: C500, which is pathogenic, and WC11, which has been attenuated by long-term maintenance in cell culture. The published genome sequence of a WC11 seed stock from a German laboratory revealed the deletion of two major regions. The sequence of a WC11 seed stock used in our laboratory also bears these deletions and, in addition, the duplication of an internal sequence in the terminal region. The larger of the two deletions has resulted in the absence of gene A7 and a large portion of gene A8. These genes are positional orthologs of the Epstein-Barr virus genes encoding envelope glycoproteins gp42 and gp350, respectively, which are involved in viral propagation and switching of cell tropism. To investigate the degree to which the absence of A7 and A8 participates in WC11 attenuation, recombinant viruses lacking these individual functions were generated in C500. Using bovine nasal turbinate and embryonic lung cell lines, increased cell-free viral propagation and impaired syncytia formation were observed in the absence of A7, whereas cell-free viral spread was inhibited in the absence of A8. Therefore, A7 appears to be involved in cell-to-cell viral spread, and A8 in viral cell-free propagation. Finally, infection of rabbits with either mutant did not induce the signs of MCF or the expansion of infected CD8+ T cells. These results demonstrate that A7 and A8 are both essential for regulating viral spread and suggest that AlHV-1 requires both genes to efficiently spread in vivo and reach CD8+ T lymphocytes and induce MCF.

of A7 in the TR. This observation led us to use the A7 STOP-207 mutant in order to avoid the risk of reversion from the duplicated region. We have also identified in Fig. 1 the 2 duplicated regions of C500 and WC11.
Line 251. Fluorescence microscopy and use of antibody 15A are not mentioned in methods Response: We have clarified these methods in the revised manuscript (lines 552-556, revised manuscript).

Line 300. is ref 43 intended? Maybe should be 35?
Response: This is a mistake. We have changed it accordingly (line 353, revised manuscript) Line 302. While it's completely acceptable to choose to study A7/A8, I think it is important that the authors mention why there was no focus on A1 in this work.

Reviewer #2:
Myster and coauthors address the genetic differences between the pathogenic C500 and an attenuated WC11 strain of the Macavirus Alcelaphine herpesvirus 1 (AlHV-1). AlHV-1 and Ovine Herpesvirus 2 induces malignant catarrhal fever in susceptible ruminants and in rabbits. In cattle and rabbits, MCF is characterized by proliferation of CD8+ T-cells, a lymphoproliferative disease with strong similarity to the pathology caused by the New World primate viruses herpesvirus saimiri and ateles in tamarins and marmosets (which is overlooked in the introduction). They focus on one of three major differences between the pathogenic and the attenuated strains, the putative glycoproteins encoded by ORFs A7 and A8, which are both lacking in the attenuated WC11 strain. Individual bacmid-technology based mutants of each of A7 and A8 were generated in the C500 strain and compared to C500 and the A7 and A8 deleted WC11 in vitro and in vivo. Although no marker rescue viruses were made, the recombinants were sequenced and the data obtained with the mutants seems consistent. However, the important rabbit experiments do not discriminate between an inability to spread efficiently in the rabbits, failure to gain access to CD8 T cell populations or an inability to reprogram such cells toward lymphoproliferation.
The authors provide an extensive comparison of the replicative behavior of C500, WC11 and two A7, one A8 mutants and mutants. Herein, depending on the cell culture system, the A7 mutants show an intermediate phenotype, increasing cell free propagation, while A8 seems to be required for cell associated spread. This may e.g. reflect alternative receptor usage for entry in these culture systems.
Finally, rabbit experiments show that the C500 A7stop, A8stop and, as expected, attenuated WC11 do not induce MCF or CD8 T cell lymphoproliferation in this experimental model. They provide evidence all strains were able to infect rabbits, by serology and detection of viral genomes. The mutants and attenuated WC11, however, were only detectable at very low levels in the spleen, and not in circulating PBMC or lymph node.

Response:
We thank the reviewer for the thorough reading of our manuscript and the relevant comments. First, we have added a reference to New World monkey herpesvirus Saimiri and Ateles (revised manuscript ref 20, lines 102-105). Second, we agree with the major comment on tte lack of mechanistic insight was missing in the original manuscript which could explain why the absence of A7 or A8 renders AlHV-1 non-pathogenic (i.e. unable to induce MCF in the rabbit model). We agree with the three main hypotheses provided by the reviewer that the lack of A7 or A8 might:

(i) impair viral spread in vivo upon infection (and thus not reach target CD8 + T cells) (ii) impair viral entry into CD8 + T cells (iii) impair AlHV-1-induced proliferation of CD8 + T cells
We now provide in the revised version of the manuscript additional data requested by the reviewer as well as unpublished data which support the hypothesis that absence of A7 or A8 affects viral spread and access to CD8 T cells (points i and ii), while being dispensable for actual reprogramming CD8 T cells (see response to major comments below). (Palmeira et al., PNAS, 2013). This important information was recently supported by yet unpublished data. Indeed, we have performed an RNA sequencing analysis on CD8 + T cells highly purified from peripheral blood of calves developing MCF and showing expansion of CD8 + T cells. From these data, we could draw a coverage map of viral RNAs expressed in CD8 + T cells throughout the genome of AlHV-1. While we could clearly detect the expression of a number of gene regions including latency-associated ORF73, we could not detect any sequencing read that would map the A7 or A8 coding sequence. This is an important information which strongly suggests that both genes are dispensable for CD8 + T cell reprogramming during MCF (as they are not expressed in infected CD8 + T cells during MCF). We are happy to share the actual data with the reviewer but would not include these confidential data in this response letter.

Regarding point (iii): we have published microarray data in 2013 showing that A7 and A8 RNA could not be detected from the lymph nodes of MCF-developing calves
Major comments: 1. Why did they not make / analyze a double A7 and A8 stop mutant, corresponding to the situation in WC11? If the double mutant would show a replicative behavior even more similar to WC11, this would be important with respect to possible influence of the other two major genetic differences, which are located on the left and right genome ends, i.e. the absent putative ORF A1 and the variable ORFs A9.5 and A10? Response: We now provide growth data of a double A7 STOP A8 STOP mutant. Although the mutant was generated at the first submission, we did not include it in the initial experiments as we believed it might confuse the main message. However, we understand that if we want to address the potential role of the genomic rearrangements and sequence divergence of strain WC11, a double impairment of A7 and A8 is of interest. Thus, we now have included the characterization of double A7 STOP A8 STOP mutant (revised Suppl. Figure S3). We have repeated the growth kinetics in BT and EBL cells (revised Fig. 3B), plaque size assay (revised Fig. 3C), viral propagation (revised Fig. 3D) and mAb 15-A (gp115 complex) immunostaining and plaque morphology (revised Fig.  4) using the A7 STOP-207 A8 STOP-159 virus. Regarding the plaque size assay in BT cells, we realized there was a mistake in the scale of the y-axis of Figure 3C, which is now adjusted and correct in the revised Figure 3. The results show that the absence of both A7 and A8 expression does not render AlHV-1 even more similar to strain WC11 in term of growth in vitro in the selected cell lines. On the contrary double A7 STOP A8 STOP mutant displayed an intermediary level of fitness which was much comparable to strain C500 BAC WT in BT cells, whereas several characteristics shared with the single A7 STOP virus essentially in EBL cells. Indeed, we could observe significantly increased cell-free virions (revised Fig. 3B, bottom), smaller plaques in EBL cells (revised Fig. 3C, bottom) and the absence of syncytia formation (revised Fig. 4 in both BT and EBL cells). The manuscript has been revised accordingly on pages 9 and 10.
Thus, there must be additional mechanism(s) than absence of A7 and A8 expression to explain why strain WC11 growth is enhanced. The obvious candidates are A1 and microRNAs identified in this region (see ref. 21 Sorel et al, 2015) and the region encompassing A9.5 to part of A10 (no data yet available on viral growth). In addition, WC11 genome has retained some part of A8 sequence which corresponds to a potential 179 aa expressed transmembrane protein. We could not detect strong mRNA expression of this sequence in cells infected with WC11 (revised Fig. 2F) but it might still be a potential explanation due to the severe genomic rearrangement found in the region of ORF50/A6/A7/A8 that might have modified viral gene regulation. Investigating this further is however out of the scope of this manuscript (lines 390-398, revised manuscript).
Finally, as both A7 and A8 expression are independently essential to induce MCF (revised Figures 6-9), we did not include in vivo experiments using the double A7 STOP A8 STOP mutant.
2. The rabbit experiments do not discriminate between an inability to spread efficiently in the rabbits, failure to gain access to CD8 T cell populations or an inability to reprogram such cells toward lymphoproliferation. There was no monitoring for viral spread (genome copies in blood?) during the infection experiment, only at endpoint. Thereby, vastly different time points are compared, and the low levels of virus genomes observed with A7 and A8 stop mutants and WC11 may reflect long term immune control of latent infection, compared to an acute disease with multiple virus genomes in the CD8 population. Response: We agree with the reviewer comment. While we had harvested PBMCs over time, we did not perform qPCR analysis to detect genome copies. The main reason was that in a rabbit (or a cow) infected with the pathogenic C500 strain, there is a long silent phase where AlHV-1 is not detectable in peripheral blood. Thus, investigating the ability of AlHV-1 to spread in rabbits is not straightforward. Nonetheless, we have extracted DNA from cryopreserved (-80°C) PBMCs from each animal over the course of the infection (in both in vivo experiments) and performed quantitative PCR to detect AlHV-1 ORF3 genomic sequence. The results are shown in revised Figures 8C and 9C and demonstrate that viral genomes are undetectable at each selected time points in animals infected with strains WC11, A7 STOP or A8 STOP . In animals infected with strain C500 WT, viral genomes could be detected from about 2 to 3 weeks post-infection, which corresponds to a time-point when persistently infected CD8 + T cells start to proliferate. In addition, we have extracted DNA from cryopreserved (-80°C) tissue biopsies of spleen, lung and liver and performed quantitative PCR to detect AlHV-1 ORF3 genomic sequence. Here, we could detect viral genomes in rabbits infected with strains WC11,