associated

22 Sepsis is the leading postnatal cause of neonatal mortality worldwide. Globally Klebsiella


INTRODUCTION
Sepsis is the leading postnatal cause of neonatal deaths, accounting for nearly one million deaths annually.(1) The incidence, causative pathogens and mortality attributable to neonatal sepsis vary globally.(2) In high-income countries, the most common causes of early-onset sepsis (EOS) are Group B streptococcus (GBS) and Escherichia coli whereas the vast majority of late-onset sepsis (LOS) cases are caused by Gram-positive organisms such as coagulase-negative staphylococci (CoNS) and Staphylococcus aureus.(3)(4)(5) In contrast, both EOS and LOS are predominantly caused by Gram-negative organisms in low-and middle-income countries (LMIC).(6)(7)(8) In Botswana and other sub-Saharan African countries, Gram-negative sepsis accounts for about 60% of neonatal sepsis, with a predominance of Klebsiella spp.(9)(10)(11) Notably, over three-quarters of the global neonatal sepsis deaths occur in LMIC signifying the critical unmet need to better understand the immune system features that predispose to, or protect against neonatal sepsis in these settings.(12) There has been a global focus on strategies to prevent neonatal infection, particularly GBS, including prophylactic antibiotics and a novel vaccine; however, geographic variation in the causes of neonatal infections highlights the need to better understand the risk factors for Gram-negative sepsis in neonates.
Since neonatal innate and adaptive immune systems are immature at birth, the immunoglobulins obtained through transplacental transfer and breastmilk provide critical protection from sepsis.(13)(14)(15)(16) Immunoglobulin G (IgG) is the primary antibody transplacentally transferred and plays an essential role in protecting neonates from infection.(17,18) Although the transfer of IgG begins at 13 weeks of gestation, it accelerates significantly after 36 weeks.Thus preterm, as compared to term, neonates are less protected by passive immunity contributing to their higher risk of sepsis.(18,19) We hypothesized that lack of maternal IgG antibodies that target specific pathogenic bacteria may predispose neonates to develop sepsis.We developed an enzyme-linked immunosorbent assay (ELISA) to evaluate the degree to which the risk of neonatal sepsis is associated with the level of IgG from maternal and neonatal serum capable of binding K.

Setting and population
This study used samples collected from a cohort of 467 neonates admitted to the neonatal unit at a tertiary referral hospital in Gaborone, Botswana from November 1, 2020 -December 31, 2021.
Participants were eligible for enrollment in the parent study if admitted to the neonatal unit within

Extraction of antibodies from dried blood spots
To extract antibodies from DBS, a 6.35-mm-diameter punch of a saturated blood spot was placed in an Eppendorf tube with 500 μl of elution buffer (phosphate buffered saline [PBS] with 0.1% BSA and protease inhibitor cocktail, Roche, catalog no.04693159001) then were incubated overnight on a shaker at 4°C.Samples were then centrifuged for 5 mins at 8,000g at 4°C to pellet debris and the supernatants were collected and stored at -20°C. .

Bacterial-binding IgG quantification for DBS eluates using ELISA
We developed an ELISA using bacteria-coated plates to determine the degree to which neonatal antibodies bind to bacterial surface epitopes of Klebsiella pneumoniae isolate MF391.This isolate was obtained from a blood culture sample in Botswana.The isolate was subcultured to a Luria broth (LB) solid media and incubated at 37°C overnight.A single colony was selected and transferred to 5 mL of LB to establish a liquid culture.The following day, the culture was centrifuged (10-minute at 4,000xg), fixed with 4% paraformaldehyde for 20 minutes and washed The remainder of the ELISA was performed as described above.The secondary anti-LPS and antiflagellin antibodies were diluted 1:5000.

Statistical analysis
Nonparametric Mann-Whitney U tests were used to compare Kleb-IgG levels and demographic characteristics between neonates with sepsis and healthy controls.A Kruskal-Wallis test was used to compare gestational age categories between the two groups.Simple linear regression was used to evaluate trends in antibody levels.We conducted an exploratory analysis using logistic regression to evaluate the association of Kleb-IgG levels with sepsis while controlling for both gestational age and days from storage to sample analysis.We conducted an additional exploratory analysis using simple linear regression to explore the association between maternal and neonatal antibody levels among neonates with and without sepsis to evaluate whether antibody transfer differed.Results were considered significant using a two-sided p value (p<0.05).
The most common organisms cultured from the neonates with sepsis were Acinetobacter spp.and Klebsiella spp.which was similar to the etiologies of sepsis in the parent study (Fig. 1).weeks but remain valid for most applications for roughly 6 months, while certain proteins can be detected after much longer periods of storage.(22)(23)(24) We, therefore, included the samples with < 4 months of storage at room temperature in this analysis.We evaluated the concentration of total IgG and Kleb-IgG from each sample since sample collection to estimate the degree of antibody degradation over time.We recovered lower IgG concentration over time from DBS samples (Fig. 2A).As such we normalized the amount of total IgG in each sample.Therefore, we did not observe a decrease in Kleb-IgG over time (Fig. 2B).Degradation of samples during room temperature storage was similar in cases compared to controls (Fig. S1).

Neonates with laboratory-confirmed sepsis have lower levels of Kleb-IgG
We next investigated whether levels of maternal and neonatal Kleb-IgG were associated with increased risk for neonatal sepsis in this cohort of 8 neonates with sepsis and 32 controls without sepsis.We found significantly lower levels of Kleb-IgG in neonates who developed sepsis compared to neonates who did not develop sepsis (p=0.012,Fig. 3A).Kleb-IgG levels were lower in neonates with sepsis regardless of etiology of sepsis (p=0.04) and not significantly different in samples from neonates with sepsis due to Klebsiella compared to other pathogens (Fig. 3B).Since neonatal IgG concentration and specificities reflect maternal serum IgG, we compared the Kleb-IgG levels in samples from mothers of neonates with and without sepsis; samples from mothers of neonates who developed sepsis tended to have less Kleb-IgG compared to mothers of controls (p=0.06,Fig. 3C).Overall, neonatal and maternal Kleb-IgG levels were correlated (r=0.27,p<0.01, data not shown).In an exploratory analysis, we unexpectedly observed that there appeared to be an inverse relationship between neonatal and maternal Kleb-IgG in dyad samples from patients with sepsis while a positive association was apparent between neonatal and maternal Kleb-IgG levels in those without sepsis.(Fig. S3). Figure 3. Kleb-IgG in neonatal and maternal samples of cases with sepsis compared to controls.A-B) Kleb-IgG binding in neonates who developed sepsis compared to matched controls.Data were normalized to standard unit, defined as the amount of Kleb-IgG in 100μl of reference serum (healthy adult) with 0.5 μg/mL IgG.Mann-Whitney U test (*p <.05).C) Kleb-IgG binding in serum from mothers of neonates who developed sepsis compared to mothers of controls.Kruskal Wallis test (*p<0.05).levels compared to neonates born at ≥37 weeks (p=0.03)(Fig. 4A).We evaluated this association in cases compared to controls and found no statistically significant difference in Kleb-IgG levels in premature neonates with sepsis compared to controls, yet a trend towards lower levels samples from neonates born at <33 weeks and between 33-37 weeks (Fig. 4B).Indeed, in an adjusted

DISCUSSION
We report an association between the level of neonatal and maternal serum IgG that bind to K. pneumoniae and all-cause laboratory-confirmed neonatal sepsis in a cohort of hospitalized neonates in Botswana.Using an ELISA to detect serum antibody reactivity to a K. pneumoniae blood culture isolate from Gaborone, Botswana, we found that neonates who developed sepsis and their mothers have lower levels of Kleb-IgG compared to controls when adjusting for gestational age and days of sample storage.K. pneumoniae was historically the predominant isolate in blood cultures at this site, (7) leading to the development of an ELISA against a K. pneumoniae isolate from this NICU.During enrollment, the epidemiology of neonatal sepsis shifted from predominantly Klebsiella spp. to a more diverse set of microbes (Fig. 1).Although there were lower levels of Kleb-IgG in preterm neonates in our study, the risk of sepsis remained significant when adjusting for GA and days since sample collection indicating that prematurity alone does not explain the association between lower Kleb-IgG levels and neonatal sepsis.Since the rate of sample degradation was similar between cases and controls, this is unlikely to account for the association (Fig. S1).In an exploratory analysis, we noted a positive linear correlation of maternal and neonatal Kleb-IgG among the whole cohort regardless of gestational age, which has been reported previously.(31) However, when we separated cases and .controls, we noted an inverse correlation in dyads with sepsis and a positive correlation in control dyads which may indicate different rates of placental transfer (Fig. S3).Many factors are known to influence transplacental antibody transfer including parity, history of infections, timing of vaccination and placental insufficiency.(14,32) Data on placental health, prior infections and maternal hypertension were not available in this cohort, thus we are unable to compare with these findings.
Our study highlights the benefits of using DBS for sample collection.Compared to traditional blood tests, DBS can be obtained via minimally invasive methods (heel stick vs. venipuncture), requires a minimal blood volume, remains stable when stored at room temperature for up to 4 months, can be transported easily with minimal risk for contamination, and is an efficient and feasible method to acquire serum samples for a variety of assays.(33,34) However, there are limitations including variation in collection volume and sample degradation which unfortunately limited the sample size in this study.

CONCLUSIONS
We propose that maternal IgG antibodies against K. pneumoniae are transferred in utero late in gestation and may be associated with a decreased risk of sepsis from diverse pathogens.
These findings suggest species-specific maternal antibodies may confer broader protection against common neonatal pathogens.This supports the ongoing development of vaccination strategies during pregnancy to increase pathogen-binding IgG levels in neonates.Further studies are needed to explore mechanisms and enhance efficient transfer of protective antibodies to premature neonates. .
three times in PBS (pH 7.4).The pellet was resuspended in PBS and diluted to achieve an optical density of 1.0.A 100 μl volume of bacteria was added to each well of a 96-well ELISA plate (Thermo Scientific Nunc, 442404).The plate was incubated overnight at 4°C then washed (0.07 M NaCl and 0.025% Tween 20 dissolved in Tris buffer) and blocked (0.007 M NaCl and 1% BSA in Tris) for 30 minutes at room temperature.DBS eluates were diluted (0.007 M NaCl, 1% BSA, 0.025% Tween 20 in Tris buffer) to a uniform IgG concentration of 0.4 ng/μl.100 μl of the diluted DBS samples were added to each well and incubated overnight at 4°C.The wells were then washed and incubated with horseradish peroxidase-conjugated anti-IgG antibody (1 mg/ml diluted 1:150,000) for an hour and developed for 15 minutes in 3,3',5,5'-tetramethylbenzidine (TMB) with 0.18M H 2 SO 4 as the stopping reagent.The absorbance at 450nm was measured immediately on an Enspire Multimode Plate Reader.A standard unit (SU) was established using reference serum from a healthy adult to establish a standard curve to normalize all samples.A standard ELISA was used to determine the IgG concentration of this serum sample.The sample was then serially diluted to create a standard curve for each experiment.We defined one SU as the optical density of 0.5 μg/mL IgG of the reference serum.Due to variability in the amount of blood collected on each card, we normalized the microbial ELISA assay by adding the same amount of IgG (i.e., 50 ng, an amount determined by titration experiments) to each well to determine the relative amount of Klebsiella-binding IgG antibodies from each sample.To determine the amount of IgG binding to common bacterial epitopes, we used E. coli lipopolysaccharide and Salmonella typhimurium flagellin.A 100 μl volume of diluted flagellin or LPS (1 μg/mL) was added to each well of a 96-well ELISA plate and incubated overnight at 4°C.

Figure 2 .Supplementary Figure 1 .
Figure 2. IgG concentration in DBS samples collected within 4 months.A) Total IgG concentration recovered from each sample as a function of time interval between sample collection and extraction.B) Relative amount of Kleb-IgG in each DBS sample, normalized to 0.4ng/µL of total IgG, as a function of time interval between sample collection and extraction.Standard Unit (SU).Simple linear regression.

Enrollment numbers, sample numbers, and causes of sepsis in overall cohort and samples included in this analysis
Caseswere matched with controls by gestational age and sample collection dates.Gestational age remained significantly different between groups, but they were otherwise similar (Table1).Three neonates were missing paired maternal samples and were excluded from maternal antibody analyses.Figure 1. .CoNS= Coagulase negative Staphyloccus spp.; MRSA= methicillin resistant Staphylococcus aureus

Table 1 . Summary of demographic variables for neonatal sepsis and control groups.
Establishing whether DBS cards stored at room temperature provide valid data is important for studies in settings where immediate storage at -20°C may be unavailable.Prior research demonstrated that most proteins on DBS cards stored at room temperature start degrading after 4 .

Kleb-IgG level by gestational age. A)
analysis, Kleb-IgG was associated with a decreased odds ratio (OR) of neonatal sepsis when adjusting for gestational age and days since sample collection (aOR 0.49, 95% CI [0.18, 0.93]).Comparison of Kleb-IgG binding in neonates stratified by gestational age.Kruskal Wallis and Mann-Whitney U test (*p<0.05).B) Comparison of Kleb-IgG binding in neonates with sepsis compared to controls, stratified by gestational age.Mann-Whitney U test.
Remarkably, neonates with sepsis due to non-Klebsiella bacteria also have lower Kleb-IgG compared to controls suggesting that Kleb-IgG might provide .cross-protectionagainst common bacterial epitopes.However, neither anti-LPS nor anti-flagellin IgG antibody levels differed between neonates with sepsis compared to controls, which argues against either being an important cross-reactive microbial epitope in this cohort.These findings suggest the possibility that cross-reactivity to other common epitopes or that multiple epitopes function together to provide protection from neonatal sepsis from a range of pathogens.By standardizing the total amount of IgG analyzed from each DBS sample, we expected to recover similar levels of Kleb-IgG across gestational ages, as we assumed that total IgG and pathogen-specific IgG would cross the placenta at the same rate.However, Kleb-IgG levels were lower in preterm neonates which may result from colonization with Klebsiella spp.later in pregnancy or increased production of anti-commensal antibodies later in gestation.Klebsiella (31)bodies may vary among pregnant people due to prior infections and differential abundance in the microbiome based on exposure, antibiotics, and other factors.(29)Similarly,Stachetalreported increased Kleb-IgG with increasing birthweight, a variable that is often colinear with gestational age.(30)Interestingly, another study focusing on pathogen-specific antibodies found no difference in anti-Klebsiella IgG by gestational age category.(31)Neither of these studies reported sepsis as an outcome.