N-dihydrogalactochitosan reduces mortality in a lethal mouse model of SARS-CoV-2

The rapid emergence and global dissemination of SARS-CoV-2 that causes COVID-19 continues to cause an unprecedented global health burden resulting in nearly 7 million deaths. While multiple vaccine countermeasures have been approved for emergency use, additional treatments are still needed due to sluggish vaccine rollout, vaccine hesitancy, and inefficient vaccine-mediated protection. Immunoadjuvant compounds delivered intranasally can guide non-specific innate immune responses during the critical early stages of viral replication, reducing morbidity and mortality. N-dihydrogalactochitosan (GC) is a novel mucoadhesive immunostimulatory polymer of β-0-4-linked N-acetylglucosamine that is solubilized by the conjugation of galactose glycans with current applications as a cancer immunotherapeutic. We tested GC as a potential countermeasure for COVID-19. GC was well-tolerated and did not produce histopathologic lesions in the mouse lung. GC administered intranasally before and after SARS-CoV-2 exposure diminished morbidity and mortality in humanized ACE2 receptor expressing mice by up to 75% and reduced infectious virus levels in the upper airway. Fluorescent labeling of GC shows that it is confined to the lumen or superficial mucosa of the nasal cavity, without involvement of adjacent or deeper tissues. Our findings demonstrate a new application for soluble immunoadjuvants such as GC for preventing disease associated with SARS-CoV-2 and may be particularly attractive to persons who are needle-averse.

. Mice were inoculated i.n. 1 day prior to the final GC 167 treatment with 10 3 or 10 4 plaque forming units (PFU) of SARS-CoV-2, a route and dose 168 intended to produce high lethality in this model. Inocula were back-titrated to verify the 169 administered dose. Tracheal swabs were collected for the first 3 days post inoculation 170 and animals were monitored daily until day 14 for weight loss and health status. that were not GC-treated rapidly declined, starting at day 4 in the 10 4 PFU group, and at 189 day 5 in the 10 3 PFU groups (Fig. 1b). Half of the GC-treated mice administered 10 4 190 PFU experienced no weight loss throughout the duration of the study (Supplementary 191 Fig . 2); conversely, their growth outpaced mock-infected counterparts treated with 192 saline alone (p<0.0001). This effect was not limited to one biological sex and did not 193 correlate with mouse starting weight at the time of study initiation. Additional mice lost 194 weight starting at day 4 but did not reach euthanasia criteria (loss of 20% of initial body 195 weight) and recovered to starting weight by day 10 post-inoculation. Relative to the 10 4 196 group, mice inoculated with 10 3 PFU SARS-CoV-2 exhibited delayed or transient weight 197 loss, with 2 mice experiencing no weight loss over the duration of the study. At the 198 higher inoculation dose of 10 4 PFU, GC significantly reduced weight loss versus 199 delivery vehicle treated controls (p<0.0001). At the lower 10 3 PFU inoculation dose, GC 200 trended toward protection from weight loss, which was confounded by non-uniform 201 disease in delivery vehicle treated controls (p=0.11). Mice treated with delivery vehicle 202 and inoculated with 10 3 or 10 4 PFU had a median survival time of 7±3.3 and 6.5±0.9 203 days, respectively (Fig. 1c). At the lower inoculation dose of 10 3 PFU, GC had an 204 efficacy of 37.5% protection from mortality (p=0.05) (Fig. 1c). At the higher inoculation 205 dose of 10 4 PFU, GC reduced SARS-CoV-2 mortality by 75% (p<0.0001) (Fig. 1d). 206 Together, these results demonstrate the potent efficacy of GC in preventing fatal SARS-207 CoV-2 disease in transgenic mice. 208 209 N-dihydrogalactochitosan reduces SARS-CoV-2 viral levels in the respiratory 210

tract. 211
We next sought to determine whether GC reduced viral levels in addition to protecting 212 mice from mortality. Infectious SARS-CoV-2 levels were assessed longitudinally in mice 213 by swabbing throats from day 1 to 3 post-inoculation (Fig. 2a). SARS-CoV-2 was 214 detectible in tracheal swabs of infected animals at day 1 post inoculation and most 215 animals had no detectible virus by day 3. Virus levels were elevated in mice inoculated 216 with 10 4 PFU versus 10 3 PFU at 1 day post inoculation (p=0.05), but differences 217 between inoculation doses were not detected after day 2 (p>0.99). GC significantly 218 reduced virus levels in tracheal swabs at day 1 and day 2 post inoculation with 10 4 PFU 219 (p=0.0005 and p=0.02, respectively). A similar effect was not observed in mice that 220 received the lower inoculum of 10 3 PFU. All delivery vehicle-treated mice had detectible 221 virus in tracheal swabs following infection, while 29.2% (7/24) of GC-treated mice had 222 no infectious virus isolated between days 1 and 3. Positive virus detection in tracheal 223 swabs was not correlated with lethal disease in mice (p=0.62, Fisher's exact test). 224 Cumulative virus levels in tracheal swabs were calculated as area under the curve for 225 individual animals (Fig. 2b). GC significantly reduced the total shed virus levels in 226 tracheal swabs in mice inoculated with 10 4 PFU from a geometric mean of 225 to 5.6 227 PFU (p<0.0001) and trended toward a reduction in animals inoculated with 10 3 PFU 228 from 138 to 15 PFU (p=0.08).   group, 2 combined experiments. LD = limit of detection, AUC = area under the curve. * 244 p < 0.05, *** p < 0.001, **** p < 0.0001. 245 As individual animals met humane experimental endpoints, mice were 246 euthanized. Infectious virus levels were measured in the lungs (Fig. 2c) and brain (Fig.  247 2d) from mice on days 6, 7, and 14. Although mean virus levels were not significantly 248 different, GC treatment trended toward reducing lung virus levels (F=2.431, p=0.06, 249 two-way ANOVA) with cumulative effects in animals inoculated with 10 3 PFU driving the 250 main treatment effect (p=0.06). 251 Comparatively high viral titers were observed in the brain at the time of death, 252 consistent with previous descriptions in this model [13]. Virus was detectible in the brain 253 in all but 2 infected animals at the time of death, indicating neuroinvasion as a likely 254 cause of morbidity. No significant differences in brain virus levels were observed 255 between treatments (F=0.8134, p=0.52, two-way ANOVA). No infectious virus was 256 detectible in the lungs or brains of animals euthanized at day 14 post inoculation. 257

N-dihydrogalactochitosan reduces the severity of histopathologic lesions 259 associated with SARS-CoV-2 infection in lungs. 260
Intranasal inoculation of hACE2 transgenic mice with PBS followed by mock inoculation 261 resulted in normal lung architecture in most mice, although several animals exhibited 262 mild alveolar septal inflammation, likely associated with i.n. administration (Fig. 3a). By 263 contrast, mice inoculated with 10 3 or 10 4 PFU SARS-CoV-2 (Fig. 3c,  to animals that did not receive GC, and inflammation was focally distributed instead of 271 widespread. To quantify histopathologic changes, lung lesion severity was scored using 272 specific criteria (Supplementary Table 2). The percent of the lung that was affected in 273 mice euthanized at 6, 7, or 14 days post inoculation as assessed by image analysis 274 software did not differ significantly between SARS-CoV-2 infected mice treated with GC 275 versus those who were not GC treated (Fig. 3d). Despite this, GC treatment significantly 276 reduced the mean lung histopathology severity score across all euthanasia days 277 (although not when the results at days 6 and 7 or 14 were analyzed separately [ Fig.  278 3e]) for mice dosed with 10 4 PFU SARS-CoV-2; scores were reduced from 4.6 to 2.5 279 (ANOVA, p < 0.0001) (Fig. 3f). Mean scores for mice treated with GC and 10 3 PFU of 280 SARS-CoV-2 trended towards being lower than for mice who did not receive GC but 281 were not significantly different (p >0.05). Together, these data show that GC reduces 282 SARS-CoV-2-induced disease in the lung of hACE2 mice, although only significantly in 283 mice administered the higher 10 4 PFU dose of SARS-CoV-2.

SARS-CoV-2 exposure. 312
We next sought to determine whether mice surviving SARS-CoV-2 inoculation 313 produced a humoral immune response that might protect them from future infection. 314 Serum from blood collected at day 14 post inoculation was assessed for neutralizing 315 antibody against the inoculation strain of SARS-CoV-2 by plaque reduction 316 neutralization test (PRNT) at the 80% neutralization threshold (Fig. 4). All mice 317 surviving virus inoculation generated neutralizing antibodies. Mice in the GC group 318 inoculated with 10 3 or 10 4 PFU generated neutralizing titers of 1:676 and 1:861 319 (geometric mean), respectively. In comparison, PBS-administered mice that received an 320 inoculation dose of 10 3 PFU had a mean neutralizing titer of 1:1448. Neutralizing 321 antibody titers were not different across treatments (p>0.99) or virus dose, although 322 PBS-administered mice inoculated with 10 4 PFU SARS-CoV-2 were unavailable for 323 comparison due to uniform lethality. These data confirm that surviving mice were 324 infected and indicate that animals were able to develop adaptive immune responses 325 that could potentially protect against reinfection. once and then euthanized 2h, 9h, or 2d later, or were treated twice (2X) and harvested 342 3d later; for these mice, the second treatment was administered 1 day before 343 euthanasia. Negative control mice were not administered FITC-GC. Histopathology was 344 evaluated and quantified using a scoring rubric developed for the nasal cavity 345 (Supplemental Table 3). Fluorescence in the nasal cavity was visualized with 346 microscopy measured by calculating the percent area of the total nasal conchae per 347 slide with a fluorescent signal. Fluorescing aggregates identified within the oral cavity of 348 several treated mice were not included in the measurements, as they were outside the 349 nasal conchae. Negative control mice had no fluorescent foci (Fig. 5a). For FITC-GC 350 treated mice, fluorescing foci were predominantly within the lumen or loosely adhered to 351 the nasal mucosa, with infrequent fluorescing areas within the epithelium (Fig. 5b-f). 352 Although small group sizes preclude statistical assessments, the fluorescent signal of 353 FITC-GC in the nasal cavity of all but 1 FITC-GC treated mouse appeared higher than 354 in negative control mice (Fig. 5g). Histologically, fluorescent material was typically 355 associated with inflammatory cells, mucous, and debris. Histology scoring in the nasal 356 concha of these mice showed that while PBS treated mice had low scores, FITC-GC 357 treatment increased scores (Fig. 5h)  viscosity, specific gravity, pH, microbiological (endotoxins and sterility), subvisible 531 particulate matter, impurities (boron, galactose, galactitol, transition metals), molecular 532 weight, and polydispersity indices. GC was presented as a 1.0% sterile solution (10 533 mg/ml) in 5 ml sealed vials and was diluted with sterile deionized water and sterile 534 filtered 20X phosphate buffered saline (PBS) to a final concentration of 0.75% GC and 535 1X PBS. Mice were anesthetized with isoflurane and 40 µL of either diluted GC or PBS 536 delivery vehicle was administered intranasally (i.n.) by a hanging drop over both nares. 537 Mice were treated identically at 3 days and 1 day prior to inoculation and 1 day post 538 inoculation. Six-week-old male (N=7) and female (N=7) mice were used to assess the 539 histopathologic effects of GC on the murine lung. Mice were GC treated on the same 540 schedule as above, and a subset were euthanized 2 or 14 days after the last GC 541 treatment. Cumulative lung lesion scores were determined using a scoring scale 542 Table 2). 543 function or rapid or depressed respiration rate. An adverse event was defined as any 555 moribund disease signs at any time over the duration of the experiment. Prior to 556 euthanasia, whole blood was collected by submandibular vein puncture under isoflurane 557 anesthesia. Whole blood was clotted for >10 min at room temperature then centrifuged 558

SARS-CoV-2 Inoculation
for 5 minutes at 8,000 x g and cleared serum was stored at -80°C. Mice were 559 euthanized by isoflurane overdose and cervical dislocation then perfused with cold 560 sterile PBS. Lung (right inferior lobe) and brain (left hemisphere) were weighed and 561 homogenized in 1-10 µL/mg DMEM with a sterile glass bead at 30 Hz for 4 minutes 562 using a TissueLyser (Qiagen, Germantown, MD) automated homogenizer. 563 Homogenates were cleared by centrifugation at 10,000 x g for 4 minutes and the 564 cleared fraction was stored at -80°C. quantitative assessment of lung inflammation, digital images were captured and 576 analyzed using ImageJ software (Fiji, NIH) to estimate the area of inflamed tissue that 577 was visible to the naked eye at subgross magnification as a percentage of the total 578 surface area of the lung section. Each lung section was scored as described 579 Table 2 ImmunoCal for 2 weeks with the solution changed every 3 days. Next they were washed 591 with PBS then processed into paraffin and embedded. The sections were taken at 4.5 592 µm every 200 µm. The stained slides were scanned at 20x on an Olympus VS120 to 593 detect the FITC signal. For the unstained slides that were scanned, they first were 594 deparaffinized, then dehydrated so they could be coverslipped and scanned, and 5 µm 595 sections were taken every 200 µm and placed on a slide for examination under an 596 epifluorescence microscope to detect the FITC signal. Nasal concha were evaluated 597 histopathologically and measurements of fluorescence were performed using ImageJ 598 (Fiji). Briefly, images were standardized using the FITC Green channel with rendering 599 set to 1500 for optimal contrast. Screenshots of each micrograph were taken, and the 600 total area of nasal cavity and mucosa was traced and recorded. Teeth, oral mucosa, 601 and connective tissues were excluded for consistency, as sections were variably cut. 602

(Supplementary
Fluorescent foci were traced and recorded. Percent area affected was calculated for 603 each slide as total fluorescent area divided by total nasal cavity area. Two sets of 604 measurements were taken per individual, with 4 total observations per treatment since 2 605 mice were included in each treatment group. antibody control. Antibody-virus dilution series were incubated for 1 hour at 37°C after 628 which they were applied to confluent Vero CCL-81 cells in single-replicate and 629 incubated for 1 hour at 5% CO2 and 37°C in a humidified incubator. Cells were overlaid, 630 incubated, fixed, and stained as described above for plaque assays. Neutralizing titer is 631 defined as the reciprocal of the dilution for which fewer than 20% of plaques were 632 detected versus the no-antibody control (>80% neutralization).   Within normal limits or rare, scattered lymphocytic infiltrates not observed in control animals but significance questionable (could be background lesion or variation of normal).
1 minimal Minimal mononuclear inflammation affecting less than 2% of the section. Inflammatory leukocyte infiltration is limited to a perivascular and/or peribronchiolar distribution with no evidence of alveolar or vascular injury.

moderate
Moderate bronchointerstitial and perivascular inflammation, increased alveolar macrophages, and/or alveolar hemorrhage/fibrin/edema (characterized as above); and/or alveolar damage (characterized by type I pneumocyte necrosis or loss with replacement by hyaline membranes, fibrin, edema, and/or necrotic debris); and/or reparative/regenerative changes (type II pneumocyte hyperplasia, atypical/multinucleated syncytial cells, or fibrosis); lesions affect 10-25% of the section.