Antiviral GHP-88310 blocks contact-mediated and airborne transmission in a ferret model of measles-like disease Authors: Carolin M Lieber, Josef D Wolf, Claire E Ruckel, Lauren A Harrison, Richard K Plemper Published: 2026 ABSTRACT AbstractThe 2025/2026 measles resurgence in North America highlighted that antivirals may be valuable in protecting child health because of mounting vaccination hesitancy. The drug candidate GHP-88310 is an orally efficacious broad-spectrum orthoparamyxovirus polymerase inhibitor, but its effect on viral transmission is unclear. Here we used canine distemper virus, which causes measles-like disease in ferrets, and established direct-contact and airborne canine distemper virus transmission models to examine pharmacological suppression of virus spread. Prophylactic GHP-88310 administration alleviated clinical signs of direct contacts and all sentinels survived. Pre- and post-exposure prophylactic GHP-88310 given twice daily to air contacts prevented transmission. Once-daily prophylactic administration mediated complete survival with all air contacts undergoing seroconversion. Therapeutic treatment of air contacts mitigated clinical signs, and animals survived, whereas all vehicle-treated air contacts succumbed. Therapeutic treatment of infected source animals shortened the contagious phase by 5 days. These results establish conceptual proof that GHP-88310 can suppress social contact transmission. Therapeutic treatment of sources promises epidemiological gain in addition to therapeutic benefit.Subject terms: Measles virus, Antivirals, Viral transmission Main In 2001, the measles initiative was launched with the objective of establishing global measles virus (MeV) control through vaccination and a possible endgame of global measles eradication1,2. Enormous progress was made towards this goal, but global measles deaths have plateaued at around 100,000 per year since 2012 without further improvement3,4. Predominantly parental vaccination hesitancy has since resulted in re-emergence of MeV in geographical areas after elimination such as the UK5, facilitated persistent transmission in large regions including central and eastern Europe6,7 and in 2025 set the stage for the largest measles outbreak in decades in the United States with over 2,280 confirmed cases, hundreds of hospitalizations and 3 deaths8. Expanding from an epicentre in West Texas, thousands of additional cases were reported in Canada and Mexico8–10. Hospitalizations and case fatalities had no apparent impact on parental risk perception of vaccination versus infection11–13. As current trends project a further decline in vaccination rates14–16, future cyclical measles outbreaks in North America are inevitable, and additional options for measles control other than traditional ring vaccination are urgently needed to improve disease management and block virus spread. MeV is an orthoparamyxovirus of the morbillivirus genus17,18 that spreads through the respiratory route. All paramyxoviruses are enveloped, single-stranded RNA viruses of negative polarity that replicate in the cytosol of infected cells, relying on the virus-encoded RNA-dependent RNA polymerase (RdRP) complex for both amplification of their RNA genomes and expression of their proteins19. Measles first presents with high fever and coryza after an incubation period of 10–14 days, followed by the characteristic maculopapular rash, conjunctivitis and diarrhoea20,21. Patients become contagious shortly after the emergence of clinical signs and continue to shed the virus until approximately 4 days after the onset of rash22. Well-behaved antivirals may contribute to addressing the measles problem through improved disease management and suppression of viral spread beyond index cases. This intended use defines key parameters of the target product profile of a viable measles drug. We require that a therapeutic is orally bioavailable for efficient administration to outpatients, is suitable for stockpiling at ambient conditions given the acyclical nature of measles outbreaks, carries a tolerability profile compatible with the high demands of a predominantly paediatric patient population and has an indication spectrum that includes an established path to clinical testing. We have previously developed two structurally distinct non-nucleoside inhibitor classes of the MeV RdRP complex, the pan-morbillivirus compound ERDRP-051923 and the broad-spectrum orthoparamyxovirus inhibitor GHP-88309, which blocks multiple clinically important pathogens in this subfamily24,25. Identified originally in a high-throughput drug screening campaign against the related orthoparamyxovirus human parainfluenza virus type 3 (HPIV3), GHP-88309 showed broad-spectrum anti-orthoparamyxovirus activity against HPIV3, a respirovirus, MeV and canine distemper virus (CDV), and emerging members of the henipavirus genus24–26. Mechanistic characterization through photo-affinity target mapping, biochemical RdRP assays and viral resistance profiling showed that the compound binds to the central cavity of the orthoparamyxovirus polymerase (L) protein and prevents RdRP initiation at the viral promoter24. Consequently, GHP class inhibitors interrupt the viral life cycle by blocking both RdRP replicase and transcriptase activity. Because MeV is a human-tropic virus without animal reservoir, we established closely related CDV25 infection of ferrets, one of its natural hosts, as a surrogate assay to study drug efficacy. Whereas CDV recapitulates all hallmarks of severe morbillivirus disease27, it exceeds MeV in pathogenicity as infected ferrets succumb invariably 10–14 days post infection (d.p.i.)25,28. We demonstrated that both ERDRP-0519 and GHP-88309 mediate complete survival and alleviate clinical signs when administered therapeutically25,28 and identified a viable therapeutic time window for treatment of morbillivirus disease with GHP-88309 (ref. 25). However, although prophylactically administered ERDRP-0519 statistically significantly prolonged time to death, animals ultimately succumbed28. The GHP-88309 chemotype was therefore selected for further development on the basis of its superior oral efficacy against morbilliviruses and its broader indication spectrum. Specifically, GHP-class compounds, but not ERDRP-0519, are orally efficacious against HPIV3 and related respiroviruses24,25, promising a viable path to clinical testing in adult volunteers as haematopoietic stem cell transplant recipients are highly vulnerable to HPIV3 infection with mortality rates of up to 75% (refs. 29,30). In recent work, we developed a clinical candidate of the GHP class, GHP-88310, that addressed an unexpected tolerability limit of original GHP-88309 in higher mammals through synthetic optimization31. In this Article, we explored the efficacy of GHP-88310 in suppressing morbillivirus transmission when used prophylactically or therapeutically. Air-contact transmission models are established for, for instance, influenza viruses and SARS-CoV-2 (refs. 32–35) to study airborne virus spread under physiological conditions but are lacking for morbilliviruses. Here, we established a robust CDV ferret airborne transmission model that revealed a strong correlation between disease dynamics and mechanism of infection. To capture both symptomatic and non-symptomatic infections, we used the emergence of clinical signs or seroconversion of immunologically naive contact animals and production of neutralizing antibodies as markers for successful transmission36. We demonstrate that pre- and post-exposure prophylactic (PREP and PEP) treatment of air contacts with GHP-88310 prevents morbillivirus transmission. Therapeutic treatment of air contacts mitigated infection, and therapeutic treatment of infected sources abbreviated the contagious phase. Results Paramyxoviruses spread through direct contact, fomites and airborne respiratory droplets and aerosols. To establish a CDV ferret morbillivirus transmission model, we explored co-housing of infected source and contact animals, recapitulating household transmission through physical contact, and shared airspace conditions, representing airborne transmission in social settings. Prophylactic GHP-88310 prevents clinical disease in direct-contact animals Sequence authentication confirmed that the CDV inoculum used in this study, recombinant CDV-5804p (ref. 37), was genetically identical to the CDV-5804p strain used previously for receptor binding site mapping38,39, ensuring binding competence to both cognate receptors relevant for efficient CDV host invasion and dissemination, canine signalling lymphocyte activation molecule (SLAM) and nectin-4 (Extended Data Table 1). We first examined direct-contact transmission, co-housing infected source animals with uninfected sentinels at a ratio of 1:3 per cage for 8 days, from 3 d.p.i. until sources succumbed to the disease (Fig. 1a). Recapitulating prophylactic treatment in a regional outbreak situation, 12 h before co-housing, each of the three contacts per cage received GHP-88309 at the highest tolerated dose of 50 mg kg−1 orally25,31, clinical candidate analogue GHP-88310 at a matching dose of 150 mg kg−1 orally based on approximately equal plasma exposure31, or vehicle volume equivalent. Treatments were continued in twice daily (bis in die, b.i.d.) regimens. We sampled all animals daily for shed virus titres in nasal lavages, determined complete blood counts (CBC) and viraemia titres in 3–4 day intervals, and assessed neutralizing antibody (nAb) titres weekly. Sampling of individual groups was discontinued when nasal lavages and peripheral blood mononuclear cells (PBMCs) were virus-negative for 3 consecutive measurements, when animals succumbed, or at 35 days after study start. Extended Data Table 1Authentication of the receptor binding domains of recCDV-5804p used in this studyH proteinNectin-4 interacting residuesSLAM interacting residuesrange of residues476–480491–495537–541524–531545–554residue of interestV478P493Y539D526/I527/S528/R529Y547/T548/R552NCBI Bioproject PRJNA1004336PHVLTYLPIQYVYDPTYDISRGDISYTHPFRLT AY386316.1 PHVLTYLPIQYVYDPTYDISRGDISYTHPFRLTThe recCDV-5804p stock used throughout this study (defined in NCBI Bioproject PRJNA1004336) fully aligns to previously reported recCDV-5804p (AY386316.1) that was used for mapping of the SLAM39 and nectin-438 receptor binding sites. Highlighted in bold are residues determining efficient binding of the viral attachment protein (H) to canine nectin-4 (left) and SLAM (right) for both recCDV-5804p preparations. Authentication of the receptor binding domains of recCDV-5804p used in this study The recCDV-5804p stock used throughout this study (defined in NCBI Bioproject PRJNA1004336) fully aligns to previously reported recCDV-5804p (AY386316.1) that was used for mapping of the SLAM39 and nectin-438 receptor binding sites. Highlighted in bold are residues determining efficient binding of the viral attachment protein (H) to canine nectin-4 (left) and SLAM (right) for both recCDV-5804p preparations. GHP-88310 prophylaxis of contact animals suppresses household transmission. a, Overview of the prophylactic assessment of twice-daily GHP-88309 or GHP-88310 to prevent direct contact transmission of CDV in ferrets. Source ferrets were intranasally (IN) infected with 2 × 105 TCID50 units of recCDV-5804p. D3, 3 d.p.i.; D11, 11 d.p.i. Contact ferrets received oral treatment, beginning 12 h before co-housing with infected source animals. b, Survival curves with statistical analysis by log-rank (Mantel–Cox) test and indicated median survival (m.s.). c,d, Progeny virus yields. Shown are PBMC-associated viraemia titre (c) and virus load in nasal lavage fluids (d), determined by TCID50 titration. e, Lymphocyte counts; normal reference range indicated by yellow shading. f, nAb titres. Monitoring of direct contact ferrets in the 150 mg kg−1 GHP-88310 b.i.d. group was discontinued 25 d.p.i. because of complete absence of clinical signs, lymphocytopaenia, and detectable viraemic and shed virus load throughout the study. Symbols in c–f represent individual animals; lines connect the geometric (in c–e) or arithmetic (in f) means. Statistical analysis with two-way ANOVA (in c–e) followed by Fisher’s least significant difference (in c) or Dunnet’s (in d and e) post hoc multiple comparisons tests. Period of co-housing is indicated by red shading (in c and d); crosses in c mark time when individual animals reached predefined end-points; LOD, limit of detection. Source data The median survival of untreated source animals was until 11 d.p.i. (Fig. 1b), which fully matched our extensive experience with this model25,31. Clinical signs emerged first at 3 d.p.i. and comprised, in order of appearance, fever, body weight loss, rash, diarrhoea and conjunctivitis (Extended Data Fig. 1a–e). Animals became viraemic 3 d.p.i. (Fig. 1c) and started to efficiently shed virus into nasal lavages 7 d.p.i. (Fig. 1d). All vehicle-treated contacts succumbed by study day 19 (Fig. 1b), indicating that transmission occurred on study days 7–8, correlating an upper respiratory CDV load of approximately 103 50% tissue culture infectious dose (TCID50) units per ml with productive spread. All contacts in the prophylactic GHP-88310 group remained disease-free during the duration of the study, and none became viraemic or virus positive in nasal lavages (Fig. 1c,d). By contrast, prophylactic GHP-88309 delayed onset, and alleviated severity, of clinical signs, but all animals of this group became viraemic and started to shed efficiently into nasal lavages, and a majority (2 of 3 animals) ultimately succumbed with a median survival of 32 days, resembling our previous experience with the structurally unrelated morbillivirus polymerase inhibitor ERDRP-0519 (ref. 28). Consistent with complete suppression of viraemia by GHP-88310, none of the animals in the group experienced any change in CBC profile, whereas animals treated with GHP-88309 suffered from severe, albeit transient, morbillivirus lymphocytopaenia, which matched in intensity that of the vehicle cohort (Fig. 1e). Two of three animals in the GHP-88310 group and the single survivor in the GHP-88309 group mounted a protective nAb response approximately 3 weeks after co-housing (Fig. 1f). Clinical signs after GHP-88310 prophylaxis of contact animals. a,b) Clinical signs. Daily body temperature (a) and weight (b) measurements. Fever (≥39.5 °C) are shown in (a) and predefined weight loss endpoint of 80% is shown in (b) as horizontal dashed line from the y-axis. Period of co-housing indicated by red shading. c-e) Clinical scores for rash (c), diarrhea (d), and conjunctivitis (e). Symbols in (a-e) represent individual animals, lines connect arithmetic means. Source data Dynamics of airborne morbillivirus transmission For long-distance transmission, we explored open air contact (oAC) and closed air contact (cAC) settings in comparison (Fig. 2a), placing source and contacts ferrets in parallel open cages or in closed cages connected through an air tube with directional flow, respectively. Contact time was 5–11 d.p.i. with a source-to-contact ratio of 1:3 in either setting. At the end of the air-exposure period, contact animals were separated into individual ventilated cages to prevent delayed secondary spread and monitored as before until study day 40. Airborne CDV transmission dynamics. a, Schematic of airborne CDV transmission set-up. Contact ferrets were co-housed with CDV-infected source ferrets in either an open- or closed-cage system from study days 5 to 11 to allow airborne exposure. All contacts received vehicle (1% methylcellulose per oral at 2 ml kg−1 b.i.d.) from study days 4.5–24. The asterisk denotes that experiments shown in a and in Fig. 3 were carried out as a double study, and the closed-cage transmission control group (green animals in a and in Fig. 3) underwent identical procedures and was applied to both study arms. b, Survival curves. Statistical analysis by log-rank (Mantel–Cox) test; m.s. is shown. c, PBMC-associated viraemia titres; crosses mark time when individual animals reached predefined end-points. d, Nasal lavage titres. Symbols in c and d represent means (source and closed-cage groups) or individual animals (open-cage group); lines connect geometric means ± geometric s.d. e, Schematic comparing clinical signs and disease progression of CDV infection in ferrets following intranasal inoculation versus airborne contact exposure, aligned to d.p.i. (intranasal) or exposure (open/closed-cage air transmission). Light coloured block indicates disease onset of the specified clinical sign of the first ferret of the group. Dark coloured blocks mark time when all animals of a group showed the respective clinical sign. Source data The median survival of animals in the cAC cohort was until study day 21 (Fig. 2b), indicating that transmission reliably occurred within the first 4–5 days of shared airspace. Viraemia and shed virus load profiles were highly consistent across all animals in the group (Fig. 2c,d), confirming synchronized transmission under these conditions. By contrast, oAC sentinels succumbed over a broad 2 week period (median survival until study day 30.5), and the onset of viraemia and virus shedding was highly variable between individual animals over a wide time window (Fig. 2b–d), attesting to overall less efficient virus spread with respect to both time to productive infection and inoculum amount transferred. A comparison of the onset of major morbillivirus-associated clinical signs after natural airborne transmission in oAC or cAC settings versus artificial intranasal inoculation illuminated a strong correlation between disease dynamics and route of infection (Fig. 2e). In each setting, all contacts were eventually infected and succumbed to the disease. However, artificially inoculated animals developed the first clinical signs (fever) 3 d.p.i., whereas cAC animals showed the first signs of disease 5 days, and oAC animals 9 days, after contact start. All individual animals in the natural infection, cAC and oAC cohorts presented with at least one disease marker at 5, 11 and 28 days after exposure, respectively. We selected the cAC setting for subsequent experiments, as it combined the natural route of infection with a predictable disease timeline. PREP and PEP GHP-88310 suppresses airborne viral transmission In addition to the vehicle control arm, we explored four GHP-88310 treatment regimens to mitigate airborne transmission (Fig. 3a): PREP at 150 mg kg−1 b.i.d.; therapeutic at 150 mg kg−1 b.i.d., started 3 days after exposure to simulate PEP; therapeutic at 150 mg kg−1 b.i.d., started therapeutically 6 days after exposure when the first clinical signs emerged; and PREP at 100 mg kg−1 once daily (quaque die, q.d.), reducing the daily dose to the previously determined lowest efficacious level after artificial inoculation31. At the end of the air-contact periods, sentinels were again transferred into individual ventilated cages. All animals in each treatment cohort survived, whereas all vehicle-treated animals succumbed with a median survival of 21 days (Fig. 3b). Before death, animals in the vehicle cohort displayed major clinical signs, whereas animals of the q.d. group showed only minor, transient disease symptoms and animals in all b.i.d. groups remained disease-free (Extended Data Fig. 2a–e). Lymphocytopaenia was severe and rapid in vehicle animals, moderate and transient in the q.d. group and fully mitigated in all animals receiving the b.i.d. regimen (Fig. 3c). Of the b.i.d.-treated animals, only those in the cohort started 6 days after air exposure developed transient mild viraemia (Fig. 3d), but the virus remained undetectable in nasal lavages (Fig. 3e). By contrast, q.d.-dosed sentinels became moderately viraemic and transiently shed virus. However, peak viraemia and shed titres remained 1–2 orders of magnitude below those of the vehicle-treated cohort. Prophylactic or therapeutic GHP-88310 suppresses productive airborne transmission. a, Study schematic of b.i.d. or q.d. prophylactic or therapeutic treatment regimens of air-contact animals. Source ferrets were intranasally infected with 2 × 105 TCID50 units of recCDV-5804p. Air contact was established 5–11 d.p.i. (grey bar). Treatments were initiated at different times relative to co-housing (coloured bars). All treatment groups were challenged intranasally with 2 × 105 TCID50 units of recCDV-5804p (yellow shading of timeline). b, Survival curves from initial challenge. c, Lymphocyte counts from initial challenge; normal reference range indicated by yellow shading. Crosses mark time when individual animals reached predefined end-points. d,e, Progeny virus yields. Shown are virus load in nasal lavages (d) and PBMC-associated viraemia titre (e). f, nAb titres in the initial and re-challenge periods. g, Survival curves after re-challenge. Statistical analyses in b and g with log-rank (Mantel–Cox) test, m.s. shown or two-way ANOVA (in c–e) with Dunnet’s post hoc multiple comparisons test. Symbols in c–f represent individual animals; lines connect geometric (in c–e) or arithmetic (in f) means. Source data Clinical signs after airborne transmission of CDV. a-c) Clinical evaluation of rash (a), diarrhea (b), and conjunctivitis (c) after initial virus challenge. d,e) Body weight (d) and temperature (e) during both initial and re-challenge. Color-coding of timeline as in Fig. 3. Weight loss endpoint of 80% (d) and fever ≥39.5 °C (e) are marked by horizontal dashed line from the y-axis. Symbols in (a-e) represent individual animals, lines connect arithmetic means. Source data We first detected CDV nAbs in animals of the q.d. group 10 days after cAC exposure was established, followed by rapid development of a robust neutralizing response (Fig. 3f). By contrast, high-dose b.i.d. treatment of cAC animals with GHP-88310 was sterilizing in the PREP and 3 days post contact (d.p.c.) PEP cohorts, indicating complete suppression of transmission, and in 1 of 3 animals of the 6 d.p.c. therapeutic group. Confirming nAb titration results, only q.d. and the two 6 d.p.c. animals with positive titre results survived a high-inoculum intranasal rechallenge with CDV 31 days after study start, whereas animals of the prophylactic and 3 d.p.c. b.i.d. cohorts and the single nAb-negative animal of the 6 d.p.i. b.i.d. group rapidly succumbed to the infection (Fig. 3g), underscoring that treatment with GHP-88310 preserved their CDV-naive immunological status. Therapeutic treatment of infected source animals alleviates clinical signs and shortens duration of virus dissemination To assess the effect of treatment of source animals on disease progression and transmission, we treated intranasally infected sources with GHP-88310 at 150 mg kg−1 b.i.d., first initiated 4 d.p.i. (Fig. 4a). In this study, sources received a reduced inoculum amount of 2 × 103 TCID50 units per animal, which we identified in pilot work as suitable to extend time to death by approximately 3 days in untreated animals (Extended Data Fig. 3), facilitating an expanded co-housing period without compromising overall outcome and disease dynamics (Extended Data Fig. 4a–h). Accordingly, direct-contact and cAC conditions were established for 5 days, from 10 to 15 d.p.i., when shed virus titres in nasal lavages of vehicle-treated source animals peaked (Fig. 4b) and viraemia titres remained high (Fig. 4c). In parallel, we initiated treatment of two additional infected animal groups 6 or 8 d.p.i., at the time of onset of fever or rash, respectively, to better appreciate the therapeutic time window of GHP-88310 (Fig. 4a). Therapeutic GHP-88310 treatment of sources shortens contagious period. a, Study schematic. Source ferrets were intranasally infected with 2 × 103 TCID50 units of recCDV-5804p. Treatment b.i.d., beginning 4 d.p.i. In 2 additional groups, treatment was initiated 6 or 8 d.p.i. Direct- or air-contact period was from 10 to 15 d.p.i. b,c, Progeny virus yields. Shown are virus load in nasal lavages (b) and PBMC-associated viraemia titre (c). Crosses in b mark time when individual animals reached predefined end-points. d, Survival curves. Analysis with log-rank (Mantel–Cox) test. e, Lymphocyte counts; normal reference range indicated by yellow shading. f, nAb titres. In b, c and e, period of co-housing is indicated by blue shading. Source and contact ferrets are shown in separate graphs in b–f for clarity. Symbols in b–f represent individual animals; lines connect geometric (in b, c and e) or arithmetic (in f) means. For vehicle source (n = 1), lines connect individual data points. Source data Correlation of viral inoculum and time-to-death. Pilot study (n = 1) testing different recCDV-5804p inoculum amounts. Survival curves analyzed with log-rank (Mantel-Cox) test. Source data Correlation of viral inoculum and clinical signs. a-h) Clinical signs and virus load in pilot study testing different inoculum amounts shown in Fig. S3. Shown are body weight (a) and temperature (b), PBMC-associated viremia (c) and nasal lavage (d) titers, lymphocyte counts (e), and rash (f), diarrhea (g), and conjunctivitis (h) scores. Symbols represent individual animals, lines connect data points. Dashed horizontal lines mark 80% body weight endpoint (a) and onset of fever ≥39.5 °C (b); lod, limit of detection. Source data All 4 and 6 d.p.i.-GHP-88310-treated animals survived the infection (Fig. 4d) and experienced only transient viraemia with minimal virus shedding into the upper respiratory tract (Fig. 4b,c). By contrast, 1 of 3 animals in the 8 d.p.i. treatment group and vehicle-treated animals succumbed to CDV at 15 d.p.i., consistent with the results of our pilot study. These results define a therapeutic time window for GHP-88310 that extends 2 days beyond the onset of first clinical signs and closes only after viraemia peak, shortly following the onset of rash. All direct and air contacts of the GHP-88310-treated sources survived, none experienced weight loss or sustained fever (Extended Data Fig. 5a,b), and none developed any clinical signs of morbillivirus disease (Extended Data Fig. 5c–e). By contrast, cACs of vehicle-treated sources died on study day 25, showing efficient transmission. Consistent with GHP-88310-mitigated viraemia, therapeutically treated animals including the 2 survivors of the 8 d.p.i. treatment group experienced only mild-moderate, transient lymphocytopaenia (Fig. 4e). Contacts of GHP-88310-treated animals remained lymphocytopaenia-free, whereas the source of the vehicle group and sentinels lost, with equivalent kinetics, >99% of circulating lymphocytes. All GHP-88310-treated animals mounted a nAb response after recovery from the primary infection (Fig. 4f). All direct- and air-contact sentinels remained immunologically naive to CDV, indicating that treatment of the source animals had suppressed transmission. Clinical signs after treatment of source animals. a-e) Measurements of body weight (a) and temperature (b) and scoring of rash (c), diarrhea (d), and conjunctivitis (e). Period of co-housing indicated by blue shading. Symbols represent individual animals, lines connect arithmetic means. Dashed horizontal lines mark 80% body weight endpoint (a) and onset of fever ≥39.5 °C (b) temperature measurements. Source data These results show that treatment initiated at onset of first clinical signs shortened, or even fully eliminated, the time during which a treated source is contagious, which may provide an epidemiological advantage in an outbreak situation in addition to direct therapeutic benefit to a treated patient. Discussion MeV lingers in room air for approximately 2 h after an infected person was present40,41. Despite high CDV shedding from source ferrets and several days of air contact, transmission between animals in open cages was inefficient. This may reflect the high air exchange rate (>14× per hour) in the vivarium, which far exceeds that of a typical indoor setting. With directional airflow between cages, airborne transmission was rapid and complete, and the ensuing disease dynamics in individual contact animals was highly reproducible. When testing prophylactic efficacy, we discovered that the original lead of the GHP class, GHP-88309, mediated complete survival only when treatment was initiated therapeutically, phenocopying our experience with structurally unrelated ERDRP-0519 (ref. 28) and potentially compromising the intended use of a measles therapeutic to preemptively treat social contacts of an index case42,43. In addition, we discovered recently that GHP-88309 has a tolerability limit of 50 mg kg−1 b.i.d. in higher mammals31. In that earlier study, we addressed the problem synthetically, obtaining favourable tolerability and pharmacokinetic profiles and establishing proof-of-concept in vivo efficacy of the resulting development candidate GHP-88310 (ref. 31). By contrast, the study presented here was motivated by the unaddressed challenge of how to efficiently contain measles outbreaks when public participation in traditional ring-vaccination campaigns is waning. Building on the identification of GHP-88310, it was the overarching objective of the work to explore whether pharmacological control of morbillivirus transmission through PREP and PEP is feasible, define the therapeutic time window for morbillivirus treatment with GHP-88310 and assess whether treatment of established morbillivirus infection may provide an epidemiological gain in addition to direct therapeutic benefit. Study results support five major conclusions: GHP-88310 provides full protection against morbillivirus disease in prophylactic and therapeutic regimen. We hypothesized that increasing the inhibitor dose will address the problem of moderate prophylactic efficacy of GHP-88309 and ERDRP-0519 (ref. 28). A 3-fold higher than standard dose of prophylactic GHP-88310 was fully efficacious, and animals remained immunologically CDV naive, confirming clearance of incoming airborne virus without adaptive immune help. Unexpectedly, however, all animals that received prophylactic GHP-88310 at 100 mg kg−1 q.d. survived, the same daily dose at which GHP-88309 failed. GHP-88310 provides full protection against morbillivirus disease in prophylactic and therapeutic regimen. We hypothesized that increasing the inhibitor dose will address the problem of moderate prophylactic efficacy of GHP-88309 and ERDRP-0519 (ref. 28). A 3-fold higher than standard dose of prophylactic GHP-88310 was fully efficacious, and animals remained immunologically CDV naive, confirming clearance of incoming airborne virus without adaptive immune help. Unexpectedly, however, all animals that received prophylactic GHP-88310 at 100 mg kg−1 q.d. survived, the same daily dose at which GHP-88309 failed. CDV-exposed contacts emerge with protective humoral immunity after prophylactic or therapeutic treatment with GHP-88310. After receiving prophylactic low-dose GHP-88310 (100 mg kg−1 q.d.), animals emerged with robust anti-CDV nAb titres at study end. If equally applicable to humans, GHP-88310-treated individuals exposed to MeV will develop robust anti-MeV immunity44,45. None of the GHP-88310-treated animals that we subsequently challenged with CDV showed signs of atypical enhanced morbillivirus disease25. CDV-exposed contacts emerge with protective humoral immunity after prophylactic or therapeutic treatment with GHP-88310. After receiving prophylactic low-dose GHP-88310 (100 mg kg−1 q.d.), animals emerged with robust anti-CDV nAb titres at study end. If equally applicable to humans, GHP-88310-treated individuals exposed to MeV will develop robust anti-MeV immunity44,45. None of the GHP-88310-treated animals that we subsequently challenged with CDV showed signs of atypical enhanced morbillivirus disease25. All PREP and PEP air contacts that received high dose (150 mg kg−1 b.i.d.) GHP-88310 remained immunologically fully naive to CDV, establishing proof of concept that a highly potent viral polymerase inhibitor is suitable to pharmacologically block viral transmission. As GHP-88310 cannot prevent host exposure to the virus, this result likely reflects both high and sustained in vivo exposure of GHP-88310 (ref. 31) and its mechanism of activity, which entails suppression of both RdRP-mediated virus replication and viral protein expression through inhibition of polymerase initiation at the promoter24. All PREP and PEP air contacts that received high dose (150 mg kg−1 b.i.d.) GHP-88310 remained immunologically fully naive to CDV, establishing proof of concept that a highly potent viral polymerase inhibitor is suitable to pharmacologically block viral transmission. As GHP-88310 cannot prevent host exposure to the virus, this result likely reflects both high and sustained in vivo exposure of GHP-88310 (ref. 31) and its mechanism of activity, which entails suppression of both RdRP-mediated virus replication and viral protein expression through inhibition of polymerase initiation at the promoter24. Therapeutic treatment of infected sources shortens the duration of virus spread. The potential epidemiological impact of antivirals is controversial. During the COVID-19 pandemic, household transmission rates were not statistically significantly impacted by treatment46. Likewise, the effect of antivirals on preventing household transmission of influenza virus is inconclusive47. However, disease dynamics of measles is slower than COVID-19 and influenza, and early clinical signs of measles such as Koplik spots48 are telling disease markers. Under these conditions, pharmacological control of MeV may provide epidemiological benefit. Therapeutic treatment of infected sources shortens the duration of virus spread. The potential epidemiological impact of antivirals is controversial. During the COVID-19 pandemic, household transmission rates were not statistically significantly impacted by treatment46. Likewise, the effect of antivirals on preventing household transmission of influenza virus is inconclusive47. However, disease dynamics of measles is slower than COVID-19 and influenza, and early clinical signs of measles such as Koplik spots48 are telling disease markers. Under these conditions, pharmacological control of MeV may provide epidemiological benefit. Well-tolerated small-molecule therapeutics are uniquely suited to meet the needs of patients with measles. Antiviral drug development for measles therapy largely focuses on two classes: small-molecule therapeutics and nAb biologics49–51. The demands of cost-effective production, high shelf stability and oral availability are best met by small molecules such as GHP-88310. Oral administration to outpatients should boost compliance compared to injectable biologics. These advantages can be offset by a low barrier to resistance52,53. Escape mutations from GHP compounds indeed emerged in cell culture24. However, all resistant viruses were attenuated in animals. GHP-88310-experienced viruses isolated from infected animals contained no resistance mutations31, suggesting that a fitness penalty may limit viral escape from GHP compounds in vivo. Future comprehensive ex vivo and in vivo resistance profiling of GHP-88310 will qualitatively and quantitatively define the genetic barrier controlling viral susceptibility to the compound. Well-tolerated small-molecule therapeutics are uniquely suited to meet the needs of patients with measles. Antiviral drug development for measles therapy largely focuses on two classes: small-molecule therapeutics and nAb biologics49–51. The demands of cost-effective production, high shelf stability and oral availability are best met by small molecules such as GHP-88310. Oral administration to outpatients should boost compliance compared to injectable biologics. These advantages can be offset by a low barrier to resistance52,53. Escape mutations from GHP compounds indeed emerged in cell culture24. However, all resistant viruses were attenuated in animals. GHP-88310-experienced viruses isolated from infected animals contained no resistance mutations31, suggesting that a fitness penalty may limit viral escape from GHP compounds in vivo. Future comprehensive ex vivo and in vivo resistance profiling of GHP-88310 will qualitatively and quantitatively define the genetic barrier controlling viral susceptibility to the compound. Direct anti-MeV efficacy of GHP-88310 in vivo is still untested. We demonstrated in a non-human primate model of measles that first-generation antiviral ERDRP-0519 was effective against a clinical MeV isolate54, establishing proof of concept that the CDV ferret data are highly representative for performance against MeV. Therapeutic time windows established for CDV in ferrets25,31 may likewise not be directly transferable to human disease, but measles will likely provide a greater opportunity for intervention because of the longer incubation and prodromal periods of MeV compared to CDV2,25,28. Other limitations of our study include currently unknown human pharmacokinetic and tolerability profiles of GHP-88310 and unknown patient compliance with treatment. Whereas consistent pharmacokinetic properties of GHP-88310, favourable tolerability in rodent and non-rodent species and consistent distribution in ferret soft-organ tissues including lung and brain55 mitigate the former concern, the question of whether guardians of paediatric patients would be open to medical intervention is real. The willingness of caretakers in the 2025 measles outbreak to follow official recommendations and administer high doses of vitamin A to their guardians56 is encouraging, as it implies that vaccination hesitancy does not equate with a general rejection of all forms of intervention. GHP-88310 combines a viable path towards clinical testing in adult volunteers through the HPIV3 indication with meeting the target product profile of a measles drug candidate for paediatric use. Prophylactic efficacy in a morbillivirus transmission model supports that GHP-88310 is suitable to suppress morbillivirus spread and contribute to outbreak control, and may aid, in conjunction with vaccination, the ultimate goal of global measles eradication. Methods Ethical compliance All animal procedures were conducted in accordance with the Guide for the Care and Use of Laboratory Animals (National Institutes of Health) and the Animal Welfare Act Code of Federal Regulations. Ferret studies were approved by the Georgia State University Institutional Animal Care and Use Committee under protocols A22035 and A25030. All experiments involving infectious agents were approved by the Georgia State University Institutional Biosafety Committee and carried out in BSL-2/ABL-2 containment facilities. Study design This study aimed to evaluate the prophylactic and therapeutic effectiveness of GHP-88310 in preventing morbillivirus transmission, using the CDV ferret model as a surrogate for MeV. Source animals were intranasally inoculated, while contact animals were exposed to the virus through direct contact transmission or airborne transmission. Viral titres in nasal lavage and PBMC samples were measured using TCID50 titration assays. Animals were randomly assigned to study groups; no data were excluded, and unless otherwise stated group sizes were 3 ferrets per condition. Figure legends state the number of biological repeats (independent experiments). Researchers were not blinded for group assignments and samples. Cells, viruses and chemicals Vero cells (ATCC CCK-81) stably expressing canine SLAM (Vero cSLAM, VDS) were cultured in Dulbecco’s modified Eagle’s medium (DMEM) with 7.5% heat-inactivated fetal bovine serum and maintained at 37 °C (5% CO2). recCDV-5804p was recovered on BSR-T7/5 cells stably expressing T7 polymerase, grown and titrated on VDS cells, and authenticated by whole genome sequencing (National Center for Biotechnology Information (NCBI) Bioproject PRJNA1004336) as reported25. GHP-88310 was synthesized and authenticated by Olon USA. Material was provided as dry powder and stored at −20 °C until reconstitution. Individual dosing stocks were dissolved in 1% methylcellulose (sonicating water bath, 37 °C) directly before administration. CDV infection of ferrets Female ferrets (Mustela putorius furo), aged 3–5 months (purchased from Triple F Farms), were allowed a 1 week rest period after arrival before being randomly assigned to groups for all experiments conducted under ABSL-2 conditions. Source ferrets were anaesthetized with a dexmedetomidine/ketamine combination and intranasally inoculated with recCDV-5804p in a total volume of 200 µl, with 100 µl administered per nare. Daily nasal lavages were performed using 1 ml of PBS containing 2× antibiotics–antimycotics (Gibco). Twice weekly, ferrets were anaesthetized with dexmedetomidine, and blood was collected from the cranial vena cava for CBC analysis, determination of nAbs and PBMC titration. CBCs were analysed using a VetScan HM5 (Abaxis). Ferrets were treated via oral gavage (Cure Catheter, 16′ sterile intermittent urinary catheter with funnel end, 12 Fr.) with either GHP-88309, GHP-88310 or vehicle-equivalent (2 ml kg−1) once or twice daily as indicated in schematics. Direct and airborne transmission conditions Direct transmission conditions were established by housing uninfected contact ferrets with a single infected source ferret in a closed single-ventilated cage (Tecniplast, GR1800 double decker). Open-cage airborne transmission conditions were created by placing three uninfected contact ferrets in an open cage (Allentown) situated in a rack and positioned 35.5 cm (front-to-front) from another cage—in a separate rack—containing an infected source ferret. Directional airflow with a rate of 2,460 ft3 min−1 was created through stationary fanning. Closed-cage transmission conditions involved placing uninfected contact ferrets in a single-ventilated cage directly connected through a 4′ ID air tube of 61 cm length with barrier grids on either end to another single-ventilated cage housing an infected ferret. Directional airflow from the cage with the infected animal to the cage with the uninfected contact was established at a rate of 85.53 in3 min−1. Relative humidity and temperature were kept consistent at 52% versus 56% and 21 °C versus 22 °C for the open- and closed-cage systems, respectively. Virus titration of nasal lavage samples and PBMCs Nasal lavage samples were clarified, and the resulting supernatants were immediately titrated using the TCID50 method on VDS cells. PBMCs were isolated from blood via chemical lysis with ACK buffer (150 mM NH4Cl, 10 mM KHCO3, 10 µM EDTA, pH 7.4), followed by two washes in PBS and resuspension in DMEM31. Viraemia titres were quantified by co-culturing serial dilutions of the extracted PBMCs with VDS cells, and results were reported as TCID50 per 106 PBMCs. nAb assay Plasma samples were heat-inactivated at 56 °C for 30 min and then clarified by centrifugation at 4,000 × g for 5 min. The clarified plasma was serially diluted twofold in serum-free DMEM, and each dilution was mixed with recCDV-5804p (2 × 103 TCID50 units). After a 75 min incubation period at 37 °C, the mixtures were transferred to VDS cells seeded in a 96-well plate, with two technical replicates per animal. Neutralization activity was assessed 3 days later by evaluating cytopathic effects via light microscopy. Statistical analysis, data preparation and sequence alignment Statistical analyses were conducted using one- and two-way analysis of variance (ANOVA), followed by appropriate post hoc multiple comparison tests as specified in figure legends. All analyses were performed using GraphPad Prism (v10.10) and/or Microsoft Excel (v16.103), and results of all statistical analyses are provided in Supplementary Data 1. Figures were generated using GraphPad Prism and assembled in Adobe Illustrator (v30.5.1). Study schematics were created in Adobe Illustrator, and animal illustrations were produced using the Adobe Illustrator vector drawing tool. The number of individual biological repeats (n values) and exact P values are indicated in the figure panels. The threshold for statistical significance was set to 0.05. MAFFT alignment was performed with Benchling via MAFFT (v7). Residues of interest identified previously38,39 were highlighted across the recCDV-5804p inoculum used in this study25 (NCBI Bioproject PRJNA1004336), which aligns to the previously reported recCDV-58024p (AY386316.1) with residue of interest bolded. Two amino acids in either direction around residue of interest are shown in Extended Data Table 1 to highlight sequence similarities. Reporting summary Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article. Supplementary information Reporting Summary Statistical analyses for Figs. 1–4. Source data Numerical raw data. Numerical raw data. Numerical raw data. Numerical raw data. Numerical raw data. Numerical raw data. Numerical raw data. Numerical raw data. Numerical raw data. Acknowledgements We thank the Georgia State University Department of Animal Resources and R. M. Cox (Georgia State University) for assistance and M. Govindarajan, M. G. Natchus and D. Kuiper (all Emory University) for coordination compound synthesis with Olon USA. The recCDV-5804p genomic plasmid, Vero-canine SLAM cells and BSR-T7/5 cells were gifts from V. von Messling (German Federal Ministry of Research, Technology and Space), Y. Yanagi (Kyushu University) and K. K. Conzelmann (formerly LMU Munich), respectively. Extended data Author contributions R.K.P. conceived the study; C.M.L. and R.K.P. designed the experiments; C.M.L., J.D.W., C.E.R., L.A.H. and R.K.P. performed the experiments; C.M.L., J.D.W. and R.K.P. curated data and prepared figures; C.M.L. and R.K.P. wrote the paper; J.D.W., C.E.R. and L.A.H. edited the paper; R.K.P. supervised the project and acquired funding. Peer review Peer review information Nature Microbiology thanks Christian Pfaller, Makoto Takeda and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Funding This study was supported, in part, by public health service grants AI071002 (to R.K.P.) and AI171403 project 3 (to R.K.P.) from the US National Institutes of Health/National Institute of Allergy and Infectious Diseases. Data availability All data that support the findings of this study are provided in this article and the extended data files. Supplementary Data 1 contains all statistical results. Sequencing data for the recCDV-5804p strain used in this study are publicly available in NCBI BioProject PRJNA1004336. Source data are provided with this paper. Competing interests C.M.L. and R.K.P. are co-inventors on a patent filing covering composition and method of use of GHP-88310 for antiviral therapy. This study could affect their personal financial status. R.K.P. reports contract testing from Enanta Pharmaceuticals and Gilead Sciences, outside of the described work. The other authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Extended data is available for this paper at https://doi.org/10.1038/s41564-026-02419-y. Supplementary information The online version contains supplementary material available at https://doi.org/10.1038/s41564-026-02419-y. References Associated Data Data Availability Statement All data that support the findings of this study are provided in this article and the extended data files. Supplementary Data 1 contains all statistical results. Sequencing data for the recCDV-5804p strain used in this study are publicly available in NCBI BioProject PRJNA1004336. Source data are provided with this paper.