Viral RNA and infectious virus in mucosal specimens from guinea pigs modelling early phases of lethal and non-lethal Lassa fever

ABSTRACT Lassa fever (LF) is endemic to broad regions of West Africa. Infection with Lassa virus (LASV), the etiologic agent of LF, results in a spectrum of clinical signs in humans, including severe and lethal hemorrhagic disease. Person-to-person transmission occurs through direct contact with body fluids or contaminated bedding and clothing. To investigate transmission risk in acute LASV infection, we evaluated viral RNA and infectious virus obtained from conjunctival, nasal, oral, genital, and rectal swab specimens from guinea pigs modelling lethal and non-lethal LF. Viral RNA and infectious virus were detected in all specimen types beginning 8 days post infection, prior to onset of fever. In the pre-clinical and clinical period, virus was isolated from a subset of nasal, oral, genital, and rectal swabs, and from all conjunctival swabs. Overall, conjunctival and nasal specimens most frequently yielded infectious virus. These findings indicate mucosal transmission risk based on virus isolation from various sites early in infection and support potential utility of minimally invasive specimen evaluation by RT-qPCR for LASV diagnostics.

Lassa virus (LASV), the etiologic agent of Lassa fever (LF), causes over 100,000 human cases annually in endemic regions of West Africa [1]. To date, mucosal shedding kinetics have not been extensively characterized in human cases or disease models, particularly during early infection. Lethal and non-lethal LF can be modelled using non-rodent-adapted clinical isolates of LASV in inbred strain 13/N guinea pigs [2,3]. Using this model, here we evaluate the presence of viral RNA and infectious virus in 5 mucosal sites at 5 timepoints early in the course of infection, representing the pre-clinical period and the period during which non-specific signs may be present, when transmission risk in humans would be highest as the index of suspicion is low. Additionally, to determine if shedding varies with disease outcome, we used two LASV strains: one that causes lethal disease in guinea pigs (Josiah) and one that causes non-lethal clinical disease (Sauerwald).
Guinea pigs that develop clinical disease after LASV infection demonstrate early non-specific signs (e.g. injected sclera, anorexia, weight loss) and elevated temperatures peaking ∼12 dpi [3]. Fever resolves in both lethal and non-lethal cases, but animals with lethal disease exhibit progressively severe weight loss and clinical signs, including respiratory insufficiency, weakness/ataxia, and marked hypothermia, necessitating euthanasia ∼17-26 dpi. To investigate viral shedding early in infection, groups of 5 animals (3 females, 2 males) infected with each LASV strain were humanely euthanized at 1 of 5 pre-determined endpoints representing: a pre-clinical period (2 and 4 dpi); onset of mild, non-specific signs (8 dpi); peak temperature elevation (12 dpi); and progression to lethal disease or resolution of non-lethal disease (16 dpi).
RNA from collected tissues (blood, eye, lung, gonad) and swabs was isolated using MagMAX Pathogen RNA/DNA Kit (RNA eluted in 75 µL; Thermo-Fisher Scientific). Genomic DNA was removed using BaseLine Zero DNase (Epicentre), and RT-qPCR was performed using SuperScript III Platinum One-Step RT-qPCR Kit (Invitrogen) with strain-specific primers and probe targeting LASV N gene. Standard curves generated by S-segment in vitro transcripts were used to quantify viral RNA (Figure 1(A); Table S1;  Table S2). No viral RNA was detected 2 and 4 dpi in swabs or associated tissues. Viremia was first detected 4 dpi at low levels in both Josiah (3 of 5) and Sauerwald (2 of 5) infected animals. Beginning 8 dpi, a subset of all swab specimen types was positive in animals from both experimental groups, and viral RNA was detected in at least two swabbed sites in each animal. Viral RNA was detected in 73 of 75 (97%) swabs collected ≥8 days from Josiah-infected animals (23 of  . Genital specimens were the only type that differed significantly by sex (female > male; p = 0.0036). In general, RNA levels were highest in genital followed by conjunctival and nasal specimens (oral and rectal were lowest). Relative RNA values and positivity coincided with peak viremia and body temperature elevation (12 dpi); levels remained stable 16 dpi with Josiah infection but declined with Sauerwald.
Here, we demonstrate viral shedding early in infection in the guinea pig model of LF, including at timepoints representing a period of non-specific signs when patients may not yet seek treatment or when LASV may not be suspected (8 and 12 dpi). Beginning 8 dpi, we detected viral RNA and isolated virus from all specimen types evaluated, including conjunctival, oral, nasal, genital, and rectal swabs. Viral RNA and virus isolation were more frequently detected and respective titres were higher in animals infected with the Josiah strain (lethal in guinea pigs); however, infectious virus was also detected in many samples from animals infected with non-lethal Sauerwald strain, indicating that shedding and transmission risk early in infection are likely independent of disease outcome.
In humans, LASV RNA has been detected in lacrimal fluid [5], saliva [5,6], pharyngeal swabs [6], seminal fluid [6][7][8], vaginal swabs [5,8] and breast milk [8]; both RNA and infectious virus have been detected in urine [6][7][8][9][10]. Our data appear consistent with these reports and build on them by investigating shedding at times points corresponding to early infection, analyzing nasal specimens, and pairing isolation attempts with viral RNA detection. All specimen types we investigated appear to be viable options for diagnostic RT-qPCR. Regarding transmission risk, conjunctival and nasal specimens yielded infectious virus most often. High titres detected in female genital swabs should be noted, along with potential underestimation of infectious virus in oral and rectal specimens due to challenges in virus isolation and titration [11,12].
Most recently, mucosal shedding and viral persistence were investigated in urine, saliva, lacrimal fluid, vaginal fluid, and seminal fluid from a large cohort of human LF survivors at discharge and serially for up to 2 years post discharge [5]. A low percentage of specimens positive in guinea pigs during acute infection were positive for viral RNA in patients at discharge (5, 9, and 21% for saliva, lacrimal, and vaginal fluid, respectively), indicating that the shedding we observed may be limited to pre-clinical and clinical periods. We detected low levels of viral RNA in only a subset of preputial swabs and were not able to isolate the virus from these samples. Thielebein et al. detected viral RNA in 80% of seminal fluid specimens at discharge [5]. The discordance in our findings is likely due to the difference in sample type (preputial swab vs. seminal fluid), resulting in underestimation of sexual transmission risk in the males we evaluated.
To our knowledge, shedding and transmission have not been reported in animal models of disease. Recently, oral, rectal and urinary shedding were characterized in Mastomys natalensis, the natural host of LASV [13]. RNA was detected in all 3 sample types; however, in contrast to our findings in guinea pigs, virus was only isolated from urine and bladder swabs, supporting different shedding profiles in the natural host compared to laboratory species developing clinical disease.
Overall, our studies indicate a high frequency of mucosal shedding and infectious virus presence early in infection in all sample types we assessed, and provide key data to consider in infection control and diagnostic assay development for LASV and other pathogenic arenaviruses.