control group (Fig. 5B). Removal (or reduction)
of NETs by systemic DNase I treatment (i.p.)
was evident in tissues of DNase I–treatedPlg−/−
mice by the significant reduction of extracellu-
lar Cit-H3 and MPO (Fig. 5C and fig. S8, A, B,
and E). Next, we generatedPlg-deficient mice
lacking neutrophil elastase (Plg−/−;Ela−/−), which
is required for NETosis ( 16 ). Deletion ofEla
abrogated NETosis in vivo (fig. S8, C to D and
F) and significantly reduced bone loss inPlg−/−
mice (Fig. 5D). Thus, neutrophil NETosis in vivo
is a mediator of fibrin-triggered oral mucosal
immunopathology.
Finally, small interfering RNAs (siRNAs) tar-
geted to deplete fibrinogen (siFbg), delivered
through lipid nanoparticles ( 17 ) produced a
reduction of circulating fibrinogen (Fig. 5E),
which was associated with significant pro-
tection from periodontal bone loss (Fig. 5F).
Thus, either inhibition of fibrin accumula-
tion or NETosis were effective in preventing
periodontal bone loss.PLGpolymorphisms are associated with
the common oral mucosal disease periodontitis
In periodontitis, local tissue inflammation and
bleeding are hallmarks of disease ( 18 ). Addi-
tionally, the infiltration of neutrophils is a
characteristic of this condition ( 19 ). Consistent
with the notion that neutrophil-fibrin inter-
actions could also mediate periodontitis in
patients, we found accumulation of extravas-
cular matrix (consistent with fibrin) (Fig. 6A)
in lesions of human periodontitis but not in
healthy tissues. Furthermore, robust neutro-
phil infiltrates were observed in the lesions of
periodontitis patients (Fig. 6A).
Human genomic data support the role of
PLG-dependent pathways in common forms
of periodontitis. Recent genome-wide associa-
tion studies have reported that polymorphisms
downstream ofPLG(rs1247559) are associ-
ated with aggressive periodontitis in Euro-
peans ( 20 , 21 ). In agreement with these previous
studies, we found thatPLGpolymorphisms
in tandem are associated with severe peri-
odontitis [PLGgene–centricP= 4.1 × 10−^4 ;
lead single-nucleotide polymorphism (SNP),
rs2465836] (Table 1 and Fig. 6B) and with
high colonization of the periodontal microbe
Aggregatibacter actinomycetemcomitans(Aa),
a pathogen known to be associated with ag-
gressive periodontitis (Aacolonization, gene-
centricP= 1.3 × 10−^3 ; lead SNP, rs4252200)
(Table 1 and Fig. 6C). Mild periodontal disease,
gingival inflammation (without periodontal
destruction), and colonization with the microbe
Porphyromonas gingivalis(Pg), a pathogen
associated with chronic forms of periodontitis,
did not yield any material evidence of asso-
ciation withPLGpolymorphisms (Table 1 and
fig. S9). These findings are consistent with pre-
vious studies of periodontitis risk–conferring
signals in the region immediately downstream
ofPLG( 20 , 21 ). Thus, genetic variation in the
PLGlocus is linked with severe and potentially
aggressive forms of human periodontitis.Silvaet al.,Science 374 , eabl5450 (2021) 24 December 2021 6 of 11
Table 1. PLG common variation is associated with common and potentially aggressive forms of periodontitis.Results of gene-centric association
testing of polymorphisms in thePLGlocus with periodontal clinical and microbiological data. Here, the healthy population serves as a control group.
Dashes indicate that no values are being reported in these sections. N/A, not applicable.Phenotypes N Gene-centricPvalue* Lead SNP [Effect allele] frequency Lead SNPPvalue............................................................................................................................................................................................................................................................................................................................................Clinical periodontal diagnosis
Healthy............................................................................................................................................................................................................................................................................................................................................ 1461 N/A (referent) –– –
Mild disease............................................................................................................................................................................................................................................................................................................................................ 812 4.3 × 10−^1 rs14224 [T] 0.60 7.6 × 10−^2
Severe gingival inflammation............................................................................................................................................................................................................................................................................................................................................ 160 5.6 × 10−^2 rs783145 [G] 0.47 2.2 × 10−^3
Severe periodontitis............................................................................................................................................................................................................................................................................................................................................† 238 4.1 × 10−^4 rs2465836 [G] 0.83 6.5 × 10−^5
............................................................................................................................................................................................................................................................................................................................................Subgingival periodontal pathogen colonization
Aa............................................................................................................................................................................................................................................................................................................................................† 989 1.3 × 10−^3 rs4252200 [A] 0.94 7.5 × 10−^5
Pg............................................................................................................................................................................................................................................................................................................................................ 989 4.1 × 10−^1 rs1806449 [T] 0.50 1.1 × 10−^1*Gene-centric associationPvalues obtained using MAGENTA (using a 110 kb upstream to 40 kb downstream window flanking PLG). †Statistically significant gene-centric association after
Bonferroni correction for multiple comparisons.Plg:
Ela:ABCPlg:+/+ -/- -/-
Fgg:051015% AreaP< 0.0001
P< 0.00010.00.20.40.6 P= 0.0007Plg-/-Ctrl DNase IMPOCit-H3390-396A/
390-396A+/390-396ABone loss (mm)24 weeks+/-
+/- +/--/-
-/--/- +/-
-/-0.00.20.60.4P= 0.0060DEPlg-/-siLuc siFbg0.00.20.40.50.30.1P= 0.0003Plasma fibrinogen (mg/mL)2.5
2.0
1.5
1.0
0.5
0.0P= 0.0003Plg-/-siLuc siFbgFBone loss (mm)Plg-/--Ctrl Plg-/--DNase ICit-H3MPOPlg+/+ Plg-/- Plg-/-Fgg390-396A/390-396APlg:+/+ -/- -/-
Fgg: 390-396A/
390-396A+/390-396A0246P< 0.0001
P= 0.0002Cit-H3 MPOBone loss (mm)Fig. 5. NETosis mediates periodontal immunopathology inPlg-deficient mice.(A) (Left) Immuno-
fluorescence staining of Cit-H3 and MPO (yellow) inPlg+/+,Plg−/−, andPlg−/−;Fgg390-396A/390-396Amouse oral
mucosal tissue sections. Scale bars, 50mm. (Right) Quantification of percentage of Cit-H3 and MPO stained
area, respectively. (BandC)Plg−/−control (Ctrl) andPlg−/−DNase I–treated mice. (B) Bone loss
quantification. (C) Staining for Cit-H3 and MPO (yellow) in gingiva. Scale bars, 50mm. (D)Plg+/−;Ela+/−,
Plg−/−;Ela+/−,Plg−/−;Ela−/−, andPlg+/−;Ela−/−mice bone loss measurements. (EandF) siRNA treatment.
(E) Plasma fibrinogen levels. (F) Bone-loss measurements ofPlg−/−-siLuc–andPlg−/−-siFbg–treated mice.
RESEARCH | RESEARCH ARTICLE