The bacterial strains used in this study were the N. meningitidis strain MC58 (WUE2135; (Mcguinness et al., 1991)) and its capsule-deficient mutant MC58siaD (WUE2425; (Ram et al., 2003)). The stock cultures were stored at -80°C in DMEM/F12 supplemented with 1% fetal calf serum (FCS), 5 mg/ml insulin and 15% glycerol. For the infection experiments, the stocks were plated on Chocolate Agar with Vitox (Oxoid, Wesel, Germany) and grown at 37°C in 5% CO₂ atmosphere overnight. Single colonies of the overnight culture were picked, dissolved and washed in phenol-red free DMEM/F12 with 1% FCS and 5 mg/ml insulin and adjusted to an optical density of 1.0 at 600nm (OD₆₀₀). The suspension, containing approximately 1x10⁹ colony forming units (CFU) per ml, was used to infect the cells and achieve a multiplicity of infection (MOI) of 100. Bacterial growth was monitored throughout all experiments.

The HIBCPP (RRID : CVCL_U566) cells were cultured in DMEM/F12 with 10% FCS and 5 mg/ml insulin. For the infection experiments, cells were grown in the inverted cell culture system as previously described (Schwerk et al., 2012). Barrier integrity as well as confluency of the cell layer was determined by measuring the transepithelial electrical resistance (TEER) by using an epithelial voltohmmeter, model Millicell- ERS STX-2 electrode system (Millipore, Schwalbach, Germany). Experiments were performed when a TEER of at least 230 Ω·cm² was reached.

Infection of the HIBCPP cells with the N. meningitidis strains MC58 and MC58siaD was described previously (Borkowski et al., 2014). In short, the confluent HIBCPP cells were grown in the inverted culture system and infected with the N. meningitidis strains at an MOI of 100 for 4h in DMEM/F12 with 1% FCS and 5 mg/ml insulin. Bacterial infection of the HIBCPP cells was followed by a wash step in PBS and subsequent extraction of whole protein lysate by using modified RIPA buffer (1x RIPA lysis buffer, 50 mM NaF, 1 mM Na₃VO₄, 1 mM PMSF, protease inhibitor cocktail). These lysates were sent to the DTU Proteomics Core facility at the Technical University of Denmark for analysis of the phosphoproteome. The infection experiment was performed five times and for each condition, protein lysates of three separate infected filters were pooled.

Whole protein content of an aliquoted amount of the lysates used for the phosphoproteomics analysis was determined using the DC Protein Assay (BioRad, München, Germany) according to the manufacturer´s instructions. 10 µg of protein was resolved by 4–12% Bis–Tris gels (Invitrogen, Karlsruhe, Germany) and electrotransferred onto nitrocellulose membranes. The MAPK p38 and its phosphorylated form p-p38 were detected using the antibodies rabbit anti-phospho-p38 and rabbit anti-p38, which were obtained from Cell Signalling (Cambridge, UK). The Immobilon Western Kit (Millipore, Schwalbach, Germany) was used to visualize immunoreactivity.

The samples were lysed using a lysis buffer (consisting of 6 M guanidinium hydrochloride, 10 mM tris(2-carboxyethyl) phosphine, 40 mM chloroacetamide, 50 mM HEPES pH 8.5). The samples were then boiled at 95°C for 5 min and subsequently sonicated on the ‘high’ setting for 5 × 30 s in a Bioruptor sonication water bath (Diagenode) at 4°C. The protein concentration was determined using a Pierce Gold BCA kit (Thermo Fisher Scientific) and 500-1000 µg were used for digestion. The samples were diluted 1:3 with 10% acetonitrile, 50 mM HEPES, pH 8.5. In the next step, LysC (mass spectrometry grade, Wako) was added in a 1:50 (enzyme to protein) ratio. Next, the samples were incubated at 37 °C for 4 h. Samples were further diluted to 1:10 using 10% acetonitrile, 50 mM HEPES pH 8.5. Next, trypsin (mass spectrometry grade, Promega) was added in a 1:100 (enzyme to protein) ratio, and samples were incubated overnight at 37 °C. The enzyme activity was quenched by adding 2% trifluoroacetic acid (TFA) to achieve a final concentration of 1%.

Prior to the Ti-IMAC enrichment, the peptides were desalted on a SOLAμ C18 plate (Thermo Fisher Scientific). For each of the samples, the C18 material was activated using 200 µl of 100% methanol (HPLC grade, Sigma), and 200 µl of buffer B (80% acetonitrile, 0.1% formic acid). The C18 bedding was subsequently equilibrated twice using 200 µl of buffer A (0.1% formic acid). Afterwards, 500-1000 µg of the sample was loaded using centrifugation at 1,000 r.p.m. The C18 bedding was washed twice with 200 µl of buffer A and the peptides were eluted into clean 1.5 ml Eppendorf tubes using 40% acetonitrile and 0.1% formic acid. The eluted peptides were concentrated in an Eppendorf Speedvac and reconstituted in 50mM HEPES (pH 8.5).

The Ti-IMAC particles were prepared according to modified manufacturers protocol. In short, the particles were agitated. 50 µl of particles per 500 µg of peptide enrichment were then placed in a 2 ml tube, which was then placed on a magnetic rack for 10 s. The shipping solution was removed and particles were washed with 200 µl 70% EtOH followed by 100 µl 1% NH₄OH and finally 3 times with 100 µl 80% Acetonitrile 1M glycoltic acid, 5% TFA, each time gently agitating for 1-5 min. Afterwards, the samples were diluted 1:1 with 80% Acetonitrile, 1M glycolic acid, 5% TFA and added to the equilibrated particles. The suspension was incubated at RT for 30 min with agitation. The bound peptides were washed 2 times with 400 µl 80% Acetonitrile, 1% TFA, followed by two wash steps in 400 µl 10% Acetonitrile, 0.1% TFA. The enriched phosphopeptides were eluted 3 times with 80 µl of NH₄OH, each time agitating for 20 min at RT. The peptides were collected in a new 0.5 ml Protein-LoBind tube and Speed-vaced for 30min at 60°C to remove most of the NH₄OH. Samples were acidified using 10% TFA and subjected to a C18 desalting step. Dried peptides were resuspended in 100 mM tetraethylammonium bromide (pH 8.5) for TMT labelling.

Labelling was done according to the manufacturer’s instructions, and labelled peptides were subsequently mixed 1:1:1:1:1:1:1:1:1:1:1:1:1:1:1:1 (16-plex), acidified to 1% TFA. The acetonitrile concentration was brought down to <5% with the use of 2% TFA. Before the mass spectrometry analysis, the peptides were fractionated using an offline Thermo Fischer Ultimate 3000 liquid chromatography system applying the high pH fractionation (5 mM ammonium bicarbonate, pH 10) at a flow rate of 5 µL min^-1. Afterwards, 20 µg of peptides were separated over a 120 min gradient (from 5% to 35% acetonitrile), with fractions being collected every 120s. The resulting 60 fractions were pooled into 30 final fractions (fractions 1 + 31, 2 + 32, 3 + 33, and so on), acidified to pH <2 with 1% TFA and loaded onto EvoSep StageTips according to the manufacturer’s protocol.

For each of the fractions, peptides were analyzed using the preset ‘30 samples per day’ method on the EvoSep One instrument. Peptides were eluted over a 44 min gradient and analyzed on an Exploris 480 instrument (Thermo Fisher Scientific) running in a data-dependent tandem mass spectrometry top-speed mode (3 s cycle time). Full mass spectrometry spectra were collected at a resolution of 60,000, with a normalized automatic gain control target of 300% and automatic maximum injection time, using a scan range of 375–1,500 m/z. The tandem mass spectrometry spectra were obtained at a resolution of 30,000 with the ‘turboTMT’ functionality enabled, with a normalized automatic gain control target of 100% and automatic maximum injection time, a normalized collision energy of 33 and an intensity threshold of 8e³. The first mass was set to 120 m/z to ensure capture of the TMT reporter ions. Precursors were isolated with a 0.7 m/z isolation window and Precursor Fit enabled to 70% at a window of 0.7 m/z. Dynamic exclusion was set to 60 s, and ions with a charge state<2, >7 and unknown were excluded. Mass spectrometry performance was verified for consistency by running complex cell lysate quality-control standards. The chromatography was monitored to check for reproducibility.

Raw files were analyzed using Proteome Discoverer 2.4. TMTPro reporter ion quantitation was enabled in the processing and consensus steps, and spectra were matched against the Homo sapiens reference proteome (UP000005640, last modified October 18, 2020, protein count 75776) obtained from Uniprot. Dynamic modifications were set as phosphorylation (S,T,Y), oxidation (M), deamidation (N,Q) and acetyl on protein N termini. Cysteine carbamidomethyl (on C residues) and TMTPro (on peptide N termini and K residues) were set as static modifications. All results were filtered to a 1% false discovery rate. TMT reporter ion quantitation done using the built-in ‘reporter ions quantifier’ node.

Prior to statistical analysis, the data was filtered so that only the phosphorylated peptides that were found in at least 75% of the samples or found in 60% in at least one sample group were included in the analysis. In this analysis the filtering resulted in a total of 3767 quantifiable peptides being used for the statistical testing. Each expression was tested for normal distribution by Shapiro-Wilk test. If a p-value > 0.05 was obtained (i.e. the expression is exhibiting a normal distribution), the groups were compared using a parametric test such as limma (Ritchie et al., 2015) which is a moderated Analysis of Variance (ANOVA) test, followed by a Benjamini Hochberg p-value correction. If the obtained p-value was below 0.05, a non-parametric statistical test was performed (Wilcoxon-Mann-Whitney test, referred to in the text as Wilcoxon test). The data is summarized in the Supplementary Tables 1–3 . A total of 11301 comparisons were tested, 2207 of which had a p-value below or equal to 0.05. When adjusting the p-value for multiple comparisons using the Benjamini-Hochberg false discovery rate procedure, 534 passed the filtration. Of these, 534 had a regulation of more than 30% and 186 have an absolute log₂ fold change higher or equal to 1. The regulations are distributed among 3 group comparisons.

Enrichment analyses were performed using the Reactome database (RRID : SCR_019316) (Yu and He, 2016). This analysis allows for the identification of cellular components, molecular function, biological processes kinases and pathways that are enriched more than anticipated by chance from a gene list based on the phosphorylated peptides. Enrichment Analyses were carried out using all regulations with an uncorrected p-value below 0.05 in the group wise statistical analysis. This analysis was carried out using the Benjamini-Hochberg Procedure and returning the enriched pathways. The adjusted p-value cut off was set at 0.05. Tables with the identified enriched pathways were constructed for all group comparisons and can be found in Supplementary Table 4 .

These enrichment analyses provide information on aspects such the cellular location where the phosphorylated peptides execute their functions, molecular activities where they are involved, and the biological processes where they might be involved. These analyses were performed using clusterProfiler (RRID : SCR_016884) (Yu et al., 2012) and DOSE (R/Bioconductor package, RRID : SCR_006442) (Yu et al., 2015). For the kinase activity enrichment analysis, the KSEA App (Wiredja et al., 2017) and Signor2 knowledge base (Licata et al., 2020) were used.

Figure 1 depicts the volcano plots of the phosphorylations that were significantly regulated according to the bioinformatic analysis of the data. The volcano plots were constructed to facilitate the visualization of the highly significant phosphorylations with larger fold changes and statistical significance of the comparisons “Con vs MC58” ( Figure 1A ) and “Con vs MC58siaD” ( Figure 1B ) as well as “MC58 vs MC58siaD” ( Figure 1C ). P-values were adjusted using the Benjamini-Hochberg procedure for limma and Wilcoxon test. The dashed horizontal line shows where the unadjusted p-value is equal to 0.05, with points below the line corresponding to p-values > 0.05 and the points above the line correspond to p-values < 0.05. The vertical dashed lines show where the log₂ fold change is equal to -1 and 1, respectively.

All statistically significant phosphorylated peptides with adjusted p-values < 0.05 and a log₂ fold changes greater than one are shown as colored dots (red and blue), representing statistically significant log₂ fold changes < -1 (blue), corresponding to phosphorylated peptides with negative regulations, or statistically significant log₂ fold changes > 1 (red), corresponding to phosphorylated peptides with positive regulations. Figures 1A, B show that the infection with the capsule-deficient mutant of N. meningitidis resulted in a higher amount of significant regulations than infection with the wildtype strain. In addition, Figure 1C confirms a high overlap of statistically relevant regulations between HIBCPP cells infected with both N. meningitidis strains as well as some differentially regulated peptides between both strains.

Figure 1D summarizes the results of the statistical analysis of the data. Tables summarizing the regulations before statistical analysis, regulations with p-values ≤0.5 and regulations with adjusted p-values ≤0.05 are shown in Supplementary Tables 1 – 3 .

Figures 2A–C show Venn diagrams comparing the impact of infection of HIBCPP cells with the wildtype strain and the capsule-deficient mutant using the regulations with an adjusted p-value <0.05. Figure 2A indicates a stronger impact of the capsule-deficient mutant on the host cell phosphoproteome than the wildtype strain. A total of 323 regulations are only identified after infection with the capsule-deficient mutant compared to 6 regulations only found after infection with the wildtype strain ( Figure 2A ). Interestingly, most regulations identified after infection with the wildtype strain can be identified after infection with the capsule-deficient mutant as well, with an overlap of 46 regulations between both strains ( Figure 2A ). Furthermore, little overlap in regulations is found when comparing control cells with infected cells (Con vs MC58 and Con vs MC58siaD) to the cells infected with the wildtype strain and capsule-deficient mutant (MC58 vs MC58siaD) ( Figures 2B, C ).

Furthermore, in a comparison of the untreated controls with the HIBCPP cells infected with MC58, 52 regulations were detected that, according to the statistical analysis, had an adjusted p-value greater than or equal to 0.05. 11 of these regulations showed a positive log₂ fold-change, while 41 regulations showed a negative fold-change and were therefore downregulated ( Figure 2D ). The comparison of the untreated controls with the HIBCPP cells infected with the capsule-deficient mutant revealed, according to statistical analysis, 369 regulations with an adjusted p-value less than or equal to 0.05. Of these, 161 were upregulated in control cells compared to infected cells while 208 were downregulated ( Figure 2D ). When both conditions were compared in which the HIBCPP cells were infected with the wildtype strain of N. meningitidis or its capsule-deficient mutant, 113 regulations were found with an adjusted p-value greater than or equal to 0.05, of which 69 were more strongly upregulated and another 44 were downregulated in cells infected with MC58 compared to cells infected with MC58siaD ( Figure 2D ).

In order to translate this primary level of information of the regulated phosphorylations into biological knowledge, enrichment analyses were carried out using all regulations with an unadjusted p-value below 0.05. The enrichment analyses were carried out as described in the Materials and Methods section. Tables summarizing the output of the analyses can be found in the Supplementary Table 4 .

Dot plots were constructed to illustrate the comparison of the annotated cellular components, molecular functions, biological processes, pathways and kinase enrichment in the different sample group tests. For each enrichment analysis, the three comparisons (Con vs MC58, Con vs MC58siaD, MC58 vs MC58siaD) showed enriched terms. The figures were made using the tools and procedures described in (Yu et al., 2012; Yu et al., 2015; Yu and He, 2016).

The waterfall plots shown in Figure 8 depict the kinase activity enrichment analysis which was created as described in (Wiredja et al., 2017). The kinases are ranked from top to bottom by the changes in the phosphorylation sites that are known to be downstream of that specific kinase from the Signor2 (Licata et al., 2020) knowledge base. Figure 8 has been filtered to only include kinases with at least 5 phosphorylated sites mapped to it from the data.

Overall, the kinase activity depicts more potentially regulated kinases after infection of HIBCPP cells with the N. meningitidis capsule-deficient mutant than after infection the with N. meningitidis wildtype strain. Of note are the kinases CDK1, CDK2 and MAPK1 which are upregulated in control cells in comparison to cells infected with the capsule-deficient mutant. Moreover, kinases found to be upregulated by the infection of HIBCPP cells with both N. meningitidis strains in this analysis (downregulated in control cells) include MAPKAPK2, RPS6KB1, RET as well as AKT1 and AKT2, whereas PRKACA activity was upregulated only by MC58siaD.

The enrichment analysis of biological processes includes the term p38 MAPK cascade and the kinase activity analysis includes a positive regulation of MAPKAPK2, which is a downstream target of the kinase p38. We therefore tested the samples that were sent to be analyzed in this study for phosphorylation of p38 for all biological replicates ( Supplementary Figure 5 ). The phosphorylation of both the MAPK p38 and Erk1/2 has previously been demonstrated in HIBCPP cells after infection with N. meningitidis (Herold et al., 2021a) and is shown here to be stronger after infection with the capsule-deficient mutant in all biological replicates.