Distinct patterns of SARS-CoV-2 transmission in two nearby communities in Wisconsin, USA

Evidence-based public health approaches that minimize the introduction and spread of new SARS-CoV-2 transmission clusters are urgently needed in the United States and other countries struggling with expanding epidemics. Here we analyze 247 full-genome SARS-CoV-2 sequences from two nearby communities in Wisconsin, USA, and find surprisingly distinct patterns of viral spread. Dane County had the 12th known introduction of SARS-CoV-2 in the United States, but this did not lead to descendant community spread. Instead, the Dane County outbreak was seeded by multiple later introductions, followed by limited community spread. In contrast, relatively few introductions in Milwaukee County led to extensive community spread. We present evidence for reduced viral spread in both counties, and limited viral transmission between counties, following the statewide “Safer at Home” public health order, which went into effect 25 March 2020. Our results suggest that early containment efforts suppressed the spread of SARS-CoV-2 within Wisconsin.


Introduction
The majority of individuals living in Dane County are White (81.5%). The next largest group 109 identifies as Hispanic or Latinx (6.3%), followed by Asian (6.0%), Black (5.9%), and American 110 Indian (0.3%) 17  . The median 116 individual in Milwaukee County is also more likely to experience poverty and to live with 117 comorbidities such as type II diabetes, hypertension, and obesity ( Table 1) (Fig 1B). 123 Sequences for this study were derived from 247 nasopharyngeal (NP) swab samples collected 124 from Dane County between 14 March 2020 through 18 April 2020, and Milwaukee County from 125 state of Wisconsin. Of these, 29 [28,31] led to introductions into Dane county whereas only 21 204 [19,21] led to introductions into Milwaukee county (Fig 4B). Surprisingly, only 9 [6, 10] of the 205 introductions into Wisconsin were associated with sequences from both counties. Furthermore, 206 these shared introductions accounted for only 20-30% of the samples from Dane and 207 Milwaukee County present in our dataset. Together, our analyses suggest that transmission 208 between Dane and Milwaukee counties has not been a principal component of viral spread 209 within either region. We find that local transmission in Milwaukee County began earlier, with an 210 introduction event in late January/early February leading to a large number of the Milwaukee 211 County sequences (Fig 4C). In comparison, most samples collected from Dane County are 212 associated with multiple introductions in late February/early March (Fig 4C). Despite the fact 213 that there were more introductions into Dane County, the reported number of cases was 214 considerably less than in Milwaukee County. This indicates that each introduction into Dane 215 County contributed less to onward viral transmission than in Milwaukee County. 216 To account for sampling bias on our estimates, we randomly sampled sequences from our set 217 of Dane and Milwaukee County samples (N = 20-240, increments of 20) and pruned all other 218 Dane and Milwaukee samples from the maximum likelihood tree. This was repeated 10 times 219 for each N, creating a set of 120 trees. We repeated the ancestral state reconstruction on each 220 of these trees and re-estimated the number of introductions (Supplemental Fig 2). The number 221 of estimated introductions into Dane County continued to increase with the number of sampled 222 sequences, indicating that these data may be undersampling the true circulating viral lineages. 223 In contrast, the number of estimated introductions into Milwaukee County decreases more 224 slowly than Dane County, consistent with a small number of introductions. Although, we cannot 225 rule out that the small number of introductions in Milwaukee County is an artifact of biased 226 sampling, where the available sequences may only represent a portion of the transmission 227 chains and not a true estimation of the total circulating viral population. Because of this, the true 228 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 10, 2020. In Dane County, estimated cumulative incidence was best predicted with the medium 245 transmission heterogeneity model based on alignment with reported incidence (Fig 5A). 246 Whereas Milwaukee County's cumulative incidence was best predicted with the model using 247 high transmission heterogeneity (Fig 5B). A greater role for superspreading events in 248 Milwaukee versus Dane County could be explained by higher population density, higher poverty 249 rates, and worse healthcare access ( is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 10, 2020. . cumulative incidence at the end of the study period (26 April)  With passive SARS-CoV-2 surveillance efforts in both counties likely missing subclinical and 258 asymptomatic SARS-CoV-2 infections, we expect the true cumulative incidence to be 259 considerably greater than the reported incidence, as has been suggested by others 39 . Indeed, 260 estimated cases were ~10x higher than reported cases in Dane County. Given that there were 261 no substantial differences in the surveillance efforts between counties, we expected more than 262 the 1. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 10, 2020. We used patterns of SARS-CoV-2 diversification in a phylodynamic model to estimate the initial 291 reproductive rate of infections in each county before official social distancing policies were 292 enacted. In this initial phase of the outbreak, the median estimated R 0 trended lower in Dane 293 County than in Milwaukee County (2.24 vs 2.82). Higher population density in Milwaukee 294 County could have contributed to a higher reproductive rate. A potential additional explanation 295 for greater community spread in Milwaukee County is that the average individual in Milwaukee 296 County, compared to Dane County, has access to fewer financial and healthcare resources and 297 is more likely to experience poverty and to live with comorbid conditions, many of which are also 298 risk factors for testing positive for SARS-CoV-2, the latter of which are also risk factors for 299 severe Coronavirus Disease (COVID-19) 16,17,41,42 . Additionally, Milwaukee County is home to a 300 higher proportion of Black and Hispanic or Latinx individuals compared to Dane County. 301 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 10, 2020. . https://doi.org/10.1101/2020.07.09.20149104 doi: medRxiv preprint Because of race-based discrimination, people belonging to these groups experience worse 302 health outcomes than White individuals, despite being treated in the same healthcare systems 303 16,17,43,44 . The social vulnerability index (SVI) is a metric ranging designed to determine how 304 resilient a community is when confronted with external stressors like natural disasters or a 305 pandemic 45 . A higher SVI indicates a community is vulnerable to experiencing worsened 306 outcomes secondary to an external stressor (range of zero to one). All of the factors mentioned 307 above contribute to a higher SVI in Milwaukee County (0.8268) compared to Dane County 308 (0.1974) 45 . While the association between SIV and SARS-CoV-2 indicidence is not significant, 309 according to a recent study, the SVI sub-components of socioeconomic and minority status are 310 both predictors of higher SARS-CoV-2 incidence and case fatality rates 46 . These sub-311 components are likely to be among the main drivers in the outbreak dynamics between Dane 312 and Milwaukee County. 313 Like most US states, in late March 2020 Wisconsin enacted a set of social distancing policies 314 aimed at reducing the spread of SARS-CoV-2. Wisconsin's order, termed "Safer at Home," was 315 enacted 25 March 2020. After this time point, the estimated R 0 was reduced by 40% or more in 316 both counties. The sequencing data is consistent with the observed reduction in positive tests, 317 as clusters expanded more slowly and new clusters arose more slowly. Throughout this time, 318 we find that the Dane County and Milwaukee County outbreaks were largely independent of one 319 another. Our data reveal only limited mixing of SARS-CoV-2 genotypes between these 320 geographically-linked communities, supporting the notion that public health policies emphasizing 321 physical distancing effectively reduce transmission between communities. Notably, "Safer at 322 Home" ended abruptly 13 May 2020, when it was overturned by the Wisconsin Supreme Court. 323 Additional sequencing and epidemiological data will be necessary to understand whether virus 324 intermingling between these counties increased after the cessation of the Executive Order. 325 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 10, 2020. . https://doi.org/10.1101/2020.07.09.20149104 doi: medRxiv preprint Viral determinants could also affect differential transmission patterns within and between Dane 326 and Milwaukee Counties. If variants with greater transmission potential exist, then early 327 introductions of such a variant into a community could contribute to greater spread there. 328 Recent reports have suggested that a point mutation in the SARS-CoV-2 spike protein-encoding 329 an aspartate-to-glycine substitution at amino acid residue 614 (D164G) may enhance 330 transmissibility. This mutation confers increased infectivity of pseudotyped murine retroviruses 331 in ACE2-expressing HEK293T cells 47 and has been proposed to be increasing in global 332 prevalence, perhaps under natural selection 48  There are some important caveats to this study. Of the total reported positives in each county 342 during the study period, high-quality sequences were available for 27% of test-positive cases in 343 Dane County, but only 5% of test-positive cases in Milwaukee County 24,25 . Despite the deep 344 sampling of SARS-CoV-2 sequences in Wisconsin relative to other regions in the US, even 345 greater targeted sequencing efforts may be required to fully capture the sequence heterogeneity 346 conferred by multiple introduction events and variable superspreading dynamics. It is possible 347 additional sequencing in Milwaukee County would uncover additional viral lineages, or that the 348 5% of cases we sequenced do not fully represent the diversity of viruses found throughout the 349 county, skewing our observations. However, in analyzing sample metadata we find no evidence 350 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 10, 2020. Through increased testing and continued sequencing efforts, it is likely that we will be able to 357 more fully understand the Milwaukee County outbreak. 358 It is also possible that other sequences from these counties relevant to our analyses were 359 collected by other groups. As of 21 June 2020, there were 477 Wisconsin sequences available, 360 but only 351 of these had geolocation information resolved to the county level. Some of the 361 remaining 126 sequences likely originated from Dane County or Milwaukee County, but we 362 cannot include these sequences in our analysis given their geolocation data resolved only to the 363 state level. Currently there is no clearly stated national-level guidance for metadata to be 364 associated with pathogen sequences. Dates and geographic locations with greater than state-365 level resolution are required to track the emergence and spread of novel pathogens like SARS-366 CoV-2. Explicit regulatory guidance from the United States enabling the disclosure of 367 sequencing data with county-level geolocation data and sampling dates would enable other 368 institutions to harmonize reporting of viral sequences and improve subsequent studies 369 comparing viral sequences from different locations. Such reporting may be especially important 370 for identifying disparities in viral transmission due to socioeconomic vulnerabilities in specific 371 counties that would otherwise be masked using state-level data reporting. 372 Here we provide the first insights into the emergence and spread of SARS-CoV-2 in Southern 373 Wisconsin. We show an early introduction of SARS-CoV-2 that did not go on to seed 374 downstream community spread. European lineages account for multiple later introductions in 375 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 10, 2020. . https://doi.org/10.1101/2020.07.09.20149104 doi: medRxiv preprint Dane County, but we find little evidence for large-scale community spread stemming from any 376 single introduction. Conversely, SARS-CoV-2 lineages from Asia account for relatively fewer 377 unique introductions into Milwaukee County and are followed by increased community spread. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 10, 2020.  is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 10, 2020. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 10, 2020. University. These trees are generated using MAFFT 53 , FastTree 54 and are available at 505 https://github.com/roblanf/sarscov2phylo/. If the closest neighbor had an ambiguous date, the 506 next closest was chosen. Any sequences which were not already in the down-sampled 507 alignment described above were added using MAFFT. IQ-TREE 55 with 1000 Ultrafast bootstrap 508 replicates 56 using the flags -nt 4 -ninit 10 -me 0.05 -bb 1000 -wbtl -czb. The 509 clock rate of the maximum likelihood tree was estimated using TreeTime 52 . We first pruned tips 510 which failed the clock filter (n_iqd = 4) and then ran TreeTime with the flags 511 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 10, 2020. . https://doi.org/10.1101/2020.07.09.20149104 doi: medRxiv preprint The number of introductions into each region was estimated using the maximum likelihood tree 512 as well as 100 of the bootstrap replicate trees. For each, we first generated a time aligned tree 513 with TreeTime with the flags infer_gtr=True max_iter=2 514 branch_length_mode='auto' resolve_polytomies=False 515 time_marginal='assign' vary_rate=0.0004 fixed_clock_rate=0.0008 57 . Tips 516 which failed the clock filter were pruned from each tree prior to running TreeTime. The 90% 517 highest posterior region was used to calculate a confidence interval for the time of each node. 518 Next, tips in the tree were assigned to either Dane County, Milwaukee County, the U.S. states, 519 or their country of origin and the ancestral states of nodes in the tree were estimated using 520 TreeTime. A sampling bias correction of 2.5 was used to account for under sampling. Nodes 521 were assigned to the region with the highest assigned probability from TreeTime. were not resolved, any Dane or Milwaukee County tips or lineages directly descending from a 534 polytomy were attributed to a single importation event -to the earliest Wisconsin node. 535 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 10, 2020. . https://doi.org/10.1101/2020.07.09.20149104 doi: medRxiv preprint We also conducted a rarefaction analysis to assess the impact of sampling within Dane and 536 Milwaukee County on the estimated number of introductions. This was done using the time 537 aligned maximum likelihood tree described above. N (20 to 240, in increments of 20) sequences 538 were randomly sampled from the set of Dane and Milwaukee County sequences and all non-539 sampled Dane and Milwaukee County sequences were pruned from the tree prior to ancestral 540 state reconstruction and estimation of the number of introductions as described above. Ten 541 replicates for each N were conducted. 542 Code to replicate this analysis is available at https://github.com/gagekmoreno/SARS-CoV-2-in- is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 10, 2020. .

737
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 10, 2020. . https://doi.org/10.1101/2020.07.09.20149104 doi: medRxiv preprint CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 10, 2020. . https://doi.org/10.1101/2020.07.09.20149104 doi: medRxiv preprint CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 10, 2020. in grey) to more effectively time-align each tree.

754
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 10, 2020. . https://doi.org/10.1101/2020.07.09.20149104 doi: medRxiv preprint . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 10, 2020. . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 10, 2020. .  is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 10, 2020.

794
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 10, 2020. . https://doi.org/10.1101/2020.07.09.20149104 doi: medRxiv preprint