In Search of Spectroscopic Signatures of Periodontitis: A SERS-Based Magnetomicrofluidic Sensor for Detection of Porphyromonas gingivalis and Aggregatibacter actinomycetemcomitans

Recently, Porphyromonas gingivalis, the keystone pathogen implicated in the development of gum disease (periodontitis), was detected in the brains of Alzheimer’s disease patients, opening up a fascinating possibility that it is also involved in the pathobiology of this neurodegenerative illness. To verify this hypothesis, an unbiased, specific, and sensitive method to detect this pathogen in biological specimens is needed. To this end, our interdisciplinary studies demonstrate that P. gingivalis can be easily identified by surface-enhanced Raman scattering (SERS). Moreover, based on SERS measurements, P. gingivalis can be distinguished from another common periodontal pathogen, Aggregatibacter actinomycetemcomitans, and also from ubiquitous oral Streptococcus spp. The results were confirmed by principal component analysis (PCA). Furthermore, we have shown that different P. gingivalis and A. actinomycetemcomitans strains can easily adsorb to silver-coated magnetic nanoparticles (Fe2O3@AgNPs). Thus, it is possible to magnetically separate investigated bacteria from other components of a specimen using the microfluidic chip. To obtain additional enhancement of the Raman signal, the NPs adsorbed to bacterial cells were magnetically attracted to the Si/Ag SERS platform. Afterward, the SERS spectra could be recorded. Such a time-saving procedure can be very helpful in rapid medical diagnostics and thus in starting the appropriate pharmacological therapy to prevent the development of periodontitis and associated comorbidities, e.g., Alzheimerʼs disease.


S2
construction, was grown aerobically at 37°C in Luria-Bertani (LB) medium and on 1.5 % agar LB plates supplemented with ampicillin at 100 μg/ml when needed.

 DNA manipulation and visualization
All basic DNA manipulation and visualization techniques as well as E. coli transformation were previously described by Sambrook et al. 1 or were conducted according to the manufacturer's instructions. Polymerase chain reactions (PCR) were carried up with Phusion™ High-Fidelity DNA Polymerase (Thermo Scientific™) under standard conditions. Oligonucleotide primer synthesis (sequences given in Table S1) and DNA sequencing were performed by Genomed, Warsaw, Poland. Restriction enzymes (Fast Digest) and T4 ligase were purchased from Thermo Scientific™. DNA was visualized in 0.8-1% agarose gels (Prona) with a use of Midori Green Advance/Direct loading dye (NIPPON Genetics).

 Mutagenesis suicide plasmid construction
Mutagenesis plasmid was constructed so that it could be used for double-crossover homologous recombination. Therefore, two -0.7 kb and 0.9 kb flanking regions on either side of the PGN1642 (PGN_RS07810) were amplified from genomic DNA of P. gingivalis 33277 by PCR with primer pairs PGN_1642_up_EcoRI / PGN_1642_up_BamHI and PGN_1642_dw_SphI / PGN_1642_dw_SalI, respectively (Table S1). Amplified fragments were subsequently cloned into pAL35 vector -a pUC19 derivative with P. gingivalis erythromycin-resistance cassette -ermF, 2 flanked with SalI/BamHI restriction enzymes. Resulting mutagenesis plasmid designated pAL53 was sequenced to confirm its correct construction.  P. gingivalis gene mutagenesis P. gingivalis deletion mutant was generated with a use of allele exchange method as previously described for other genes; 3 pAL53 suicide plasmid was introduced by electroporation into electrocompetent wild-type P. gingivalis cells prepared according to protocol included in. 4 Obtained clones were selected on erythromycin plates and double-crossover genomic recombination was confirmed by PCR specific for mutated region ( Figure S1a, b, Table S1) and DNA sequencing of pertinent manipulated region.

 Exclusion of mutation polar effect
Analysis of P. gingivalis 33277 genome fragment encoding PGN_1642 revealed that the gene of interest may be a part of larger transcriptional unit PGN_1643-1642-1641-1640. Therefore, we checked if replacement of PGN_1642 gene with erythromycin cassette (ermF) didn't alter transcription (polar effect) of two following genes PGN_1641 and PGN_1640. Total mRNA was isolated with TRIzol Reagent (Thermo Scientific) from P. gingivalis AL022 mutant and reverse-transcribed with SuperScript III First-Strand Synthesis System (Invitrogen) using universal primer provided. Resulting cDNA was then used as a template for PCR reactions detecting the presence of transcripts for PGN_1641 and PGN_1640 genes ( Figure S1a, c). Experiment confirmed no polar effect of introduced mutation in AL022 strain. Transcripts for both genes (PGN_1641 and PGN_1640) were present in AL022 strain (RT+) and absent in reactions performed without reverse transcriptase (RT-; negative control) shown on Figure S1c.
Genomic DNA (g) of AL022 strain was used as positive control template. Primers utilized for this experiment are listed in Table S1 and marked on a scheme ( Figure S1a). ATCAAGCAGGGACGGAAG None * restriction site sequence was underlined; not all sites used for cloning purposes # also used for checking mutation polar effect  Limit of detection (LOD) Figure S7. The limit of detection (LOD) of bacterial cells for the proposed SERS-microfluidicbased method. In order to verify the LOD, 1 ml of each of the following concentrations of A. actinomycetemcomitans 652: 10 5 , 10 4 , 10 3 , and 10 2 cfu/ml (red spectra) was placed in the microfluidic device. The samples were prepared by serial dilution of the initial sample which was adjusted for 0.5 McFarland turbidity standards (10 8 colony forming units/ml). 5 The spectrum of A. actinomycetemcomitans 652 (dark yellow), collected during standard measurement and described in the main manuscript, was given for comparison.