Plasma Levels of Mannan-Binding Lectinin Relation to Periodontitis and SmokingGaia Maffei,* Nannette Brouwer,† Koert M. Dolman,† Ubele van der Velden,* Dirk Roos,†

and Bruno G. Loos*

Background: Mannan-binding lectin (MBL) is an importantmolecule of innate immunity; it acts as an opsonin and stimu-lates the classical complement pathway. Moreover, it hasbeen suggested that MBL acts as a weak acute phase protein.We investigated whether MBL levels are increased in peri-odontitis, and we tested whether individuals deficient forMBL are more susceptible to periodontitis.

Methods: A total of 219 subjects participated in the study.Plasma samples from 115 periodontitis patients and 104healthy controls were taken, and the MBL levels were mea-sured by enzyme-linked immunosorbent assay (ELISA). MBLlevels were analyzed in relation to periodontitis, taking into con-sideration age, gender, ethnic and educational background,and smoking status. In some analyses, subjects with MBLplasma levels <0.8 mg/ml were considered MBL deficient.

Results: MBL plasma concentrations were not significantlydifferent in moderate and severe periodontitis comparedto controls (1.6, 1.4, and 1.6 mg/ml, respectively). Also, theprevalence of MBL deficiency was not found to be differentbetween controls and moderate and severe periodontitis(45%, 37%, and 36%). However, among all subjects andamong the non-deficient subjects, MBL levels were markedlyincreased in heavy smokers (>10 cigarettes per day), irre-spective of periodontal disease status, in comparison to non-smokers and light smokers. MBL plasma levels did not showa correlation with plasma C-reactive protein (CRP) and werealso not related to the prevalence of specific periodontalpathogens.

Conclusions: MBL levels were not elevated in periodontitis,and MBL deficiency was not related to susceptibility for peri-odontitis. The fact that MBL levels were higher among heavysmokers is the subject of further investigation. J Periodontol2005;76:1881-1889.

KEY WORDS

Complement; innate immunity; mannan-binding lectin;periodontitis; smoking.

Periodontitis is a chronic inflam-matory disease affecting the sup-porting tissues of the teeth. A

primary risk factor for periodontitis isdental plaque.1 Specifically, a numberof Gram-negative bacteria, includingActinobacillus actinomycetemcomitans,Porphyromonas gingivalis, and Tanner-ella forsythensis, have been implicatedas periodontal pathogens.2 The peri-odontal pathogens elicit chronic inflam-matory reactions and formation ofgranulation tissue at the cost of peri-odontal ligament and alveolar bone. Ifleft untreated, teeth loosen and eventu-ally exfoliate. Epidemiological reportscurrently estimate that 10% of the pop-ulation develops severe forms of peri-odontitis, while moderate forms of thedisease occur in 40% of the popula-tion.3-5

It is generally accepted that the pres-ence of periodontal pathogens in the den-tal plaque is necessary but not sufficientfor periodontitis to develop. Host immu-nity and genetic and environmental fac-tors determine part of the susceptibilityto and severity of periodontitis.6,7 For ex-ample, smoking is regarded as a majorenvironmental factor associated with pe-riodontitis.8-11 The unfavorable effects ofsmoking on the periodontal parametersmay be related to local ischemic eventsand to functional changes in parts ofthe immune system.12-14

Mannan-binding lectin (MBL) is an im-portant component of innate immu-nity,15,16 and it may play a role in thedefense against invading microorgan-isms in periodontitis. MBL is produced

* Department of Periodontology, Academic Center for Dentistry Amsterdam, University ofAmsterdam and Vrije University, Amsterdam, The Netherlands.

† Department of Experimental Immunohematology, Sanquin Research and LandsteinerLaboratory, Academic Medical Center, University of Amsterdam.

J Periodontol • November 2005

1881

in the liver, and it belongs to the family of collectins.Collectins contain a collagen-like region and a carbo-hydrate-binding lectin domain. In this way, MBL canrecognize carbohydrate structures, specifically theterminal mannose groups on the surface of a varietyof microorganisms. These include Neisseria menin-gitidis, Candida species, Aspergillus fumigatus,Staphylococcus aureus, beta-hemolytis group Astreptococci, and anaerobic bacteria like Bifido-bacterium bifidum and Veillonella dispar.17-19

Interestingly, the periodontal pathogens A. actinomy-cetemcomitans and P. gingivalis also appear to havemannan-containing polysaccharides on the cell sur-faces.20-24 For example, the predominant serotypeantigen of A. actinomycetemcomitans resides in amannan-rich polysaccharide.21,24 When ligated tobacteria, MBL is able to interact with serum-derivedserine proteinases known as MBL-associated serineproteinases (MASPs). This interaction results in theformation of the MBL complex.25 The interaction ofMBL with MASPs is analogous to the interaction ofthe complement factor C1q with complement factorsC1s and C1r in the initial steps of the classical com-plement activation pathway. MBL complexed withMASPs and the carbohydrate patterns on bacteriainteracts with C4, and in this way MBL activatesthe classical complement pathway in an antibody-independent manner (MBL pathway).16 In addition,it has been suggested that MBL can act directly asan opsonin.26

Plasma levels of MBL are on average 1.2 to 1.6mg/ml in white populations.27,28 Subjects with levelsbelow 0.5 to 1.00 mg/ml are considered MBL defi-cient.28 In our laboratories, for patient diagnosticand research purposes, we use 0.8 mg/ml as cutoffvalue for deficiency. Impaired serum levels of MBLhave been associated with severe infections in adultsand children: recurrent lung infections, recurrent otitismedia, diarrhea, and septicemia.29-31 On the otherhand, other studies have found that the frequency ofMBL deficiency in patients with signs of infectionsdid not differ from that of controls.32,33

MBL is also regarded as an acute-phase protein.MBL plasma levels have been reported to increaseduring infections and inflammatory processes. Forexample, the circulating level of MBL was found to in-crease 1.5- to 3-fold in patients undergoing surgery orsuffering from malaria.32,34 It has been shown that thepromoter region of the MBL gene contains acute-phase response elements.35 However, in comparisonwith C-reactive protein (CRP), the prototypical acute-phase protein, the increase in MBL levels is modest,and therefore MBL has been designated as a weakacute-phase reactant.16,36

The purpose of the present study is to investigatewhether MBL plasma levels are increased in periodon-

titis patients and to elucidate whether MBL deficiencyis associated with periodontitis.

MATERIALS AND METHODS

Study PopulationAtotal of 219 subjects participated in this study.Thesesubjects were all adults mainly of white background(189; 86%). The study population included a consec-utiveseriesof115patientswhowerereferredto theDe-partment of Periodontology at the Academic Centerfor Dentistry Amsterdam (ACTA), Amsterdam, TheNetherlands, for diagnosis and treatment of periodon-titis (recruitment period 1995 to 1996). All patientswere initially screened in our ‘‘new patient intake cli-nic’’; theywerepreviouslyuntreatedandwererequiredto have a Community Periodontal Index of TreatmentNeeds(CPITN)scoreof four inat leastonesextantcon-comitant with radiographic evidence for alveolar bonedestruction. Furthermore, the study population in-cluded 104 control subjects who were registered forrestorative dental procedures or regular dental check-ups (once or twice per year; recruitment period: 1996to 1997 for N = 81 and 2001 for N = 23). Only con-trol subjects with less than or equal to one tooth perquadrant missing (third molars excluded) and a radio-graphicdistance£3mmbetweenthecemento-enameljunction and the alveolar bone crest at all approximalsites on bite-wings £1 year old were included in thestudy. From the dental charts, it was ensured that nopathologically deepened pockets were present. How-ever, control subjects may have had gingivitis. Fromprevious studies,37-40 with the same control subjects,it was clear that the controls had a significantly differ-ent subgingivalmicrofloraandblood(plasma)param-eters. Informed consent was obtained from all studysubjects. This study was approved by the Medical Eth-ical Committee of our institution.

A medical history was taken by a written question-naire and interview. The study subjects did not includepregnant women and individuals who suffered, apartfrom periodontitis, from any given systemic chronicmedical condition, for example diabetes or rheuma-toid arthritis. For all participants, smoking habits wererecorded and subjects were classified as light or heavysmokers (respectively, £10 cigarettes per day [cig/day] or >10 cig/day) or as non-smokers. Subjectswho had stopped smoking were also classified asnon-smokers. Educational level less than, equal to,or greater than high school was recorded and usedas a surrogate marker for socioeconomic status.

Since attachment level measurements were notavailable (only probing depth measurements are partof the routine in our ‘‘new patient intake clinic’’), weused radiographs to estimate the severity of periodon-tal destruction. For each periodontitis patient, a setof full-mouth radiographs was available and the

Mannan-Binding Lectin in Relation to Periodontitis and Smoking Volume 76 • Number 11

1882

individual number of teeth present was determined.For each patient, all teeth were radiographically ex-amined on mesial and distal aspects. The number ofteeth without bone loss, with bone loss extending be-yond one-third of the root length, and teeth with boneloss ‡50% of the root length was determined. Patientswith greater than or equal to seven teeth with ‡50%bone loss were classified as having severe periodon-titis, while the remainder of patients were classifiedas having mild to moderate periodontitis.38,40

Venous Blood SamplesVenous blood samples were obtained by venepunc-ture without excessive venous stasis. The blood sam-ples were collected in EDTA-containing vacuumtubes.‡ The tubes were then centrifuged at 1500 · gduring 10 minutes. The obtained plasma sampleswere stored in aliquots at -80�C until use.

MBL DeterminationThe plasma levels of functional MBL were determinedby enzyme-linked immunosorbent assay (ELISA). Allincubation steps were performed with 100 ml per well,and between every incubation step the plates werewashed five times with demineralized water. Plates§

were coated with coating buffer (0.1 M NaHCO3, pH9.6) containing 50 mg/ml mannani and incubated over-night at room temperature.

Sample preparation. Sample dilutions were madein TTG/Ca2+ (20 mM Tris pH 7.4, 150 mM NaCl,0.02% [w/v] Tween-20, 0.2% [w/v] gelatin, 10 mMCaCl2) with 10 mg/ml heparin. Each plasma samplewas tested in duplo in four dilutions: 1:3, 1:9, 1:27,and 1:81. The MBL concentration for each samplewas calculated as a mean of a duplo in the linear partof the standard curve. Standard curves were madefrom pooled plasma, one dilution 1:3 followed by eightdilutions 1:2.

Incubation steps. Sample dilutions were incubatedin the mannan-coated plates for at least 1 hour, shak-ing, at room temperature. After washing, the plateswere incubated with 20 mg/ml biotinylated anti-MBL-antibody¶ in TTG/Ca2+, for 1 hour, shaking at roomtemperature. After washing, streptavidin-poly horse-radish peroxidase (pHRP) 1:10,000 diluted in Tris/Ca2+, and 2% milk was added and incubated for 20to 30 min, shaking at room temperature. Fresh sub-strate solution was prepared (0.11 M CH3COONapH 5.5, 100 mg/ml 3,39,5,59-tetramethylbenzidine[TMB],# 0.033% [w/v] H2O2), and, after washing, theplates were incubated until the blue color clearlyshowed, while the H2O controls were still blank. Stopsolution (2 M H2SO4) was added to the wells, and theresulting yellow color intensities were read at 450 nmin an ELISA plate reader.** The minimal detectionlimit of the analysis was 0.05 mg/ml.

Available Data Set and Methodologies UsedFrom two previous studies,11,37 with the majority ofthe same subjects, we used data of plasma levels ofcotinine and CRP, which were available for all patients(N = 115) and for 75% of the controls (N = 78). Plasmacotinine was determined by ELISA according to themanufacturer’s instructions.†† Briefly, a standard in-hibition curve was obtained, and plasma sampleswere tested undiluted at 1:10 and 1:20 dilutions.Plasma levels of CRP were determined using a latexenhanced nephelometric method, with commerciallyavailable reagents and instruments according tothe manufacturer’s instructions.‡‡ In addition, dataon the prevalence of periodontal pathogens wereavailable for all patients (N = 115).39 In brief, pooledsubgingival samples were taken from the deepestbleeding site per quadrant to assess the prevalenceand the proportions of the following bacteria: A. acti-nomycetemcomitans, P. gingivalis, T. forsythensis,Prevotella intermedia, Fusobacterium nucleatum,Peptostreptococcus micros, and Campylobacter rec-tus. The sampling, culturing, and identification ofthe samples were performed according to previouslydescribed standard procedures.41 The detection levelfor A. actinomycetemcomitans is 10 cells/ml and, forall other target species, 104 cells/ml.

Statistical AnalysisStatistical analyses and box plot generation wereperformed.§§ Differences among subject groups inbackground characteristics were analyzed withMantel-Haenszel tests. For presentation of data andstatistical purposes, MBL values in subjects withELISA results below the detection limit (N = 7) wereassigned values of half the lower limit of detection.42

Differences between the three study groups (controlsubjects and moderate and severe periodontitis pa-tients) for MBL levels were tested with analysis of co-variance (ANCOVA), taking also into considerationthe following potential confounding factors: educa-tional level and smoking status. Moreover, the follow-ing potential confounding covariates were entered inthe model: age, gender, and ethnicity. Whenever theoverall ANCOVA model between three groups wassignificant, post-hoc testing between any two groupswas performed similarly. In these latter analyses, Pvalues underwent Bonferroni corrections. Similarly,MBL levels among three groups of subjects basedon smoking habits were analyzed. Partial correlation

‡ Venoject II, Terumo Europe BV, Leuven, Belgium.§ Micro-well plates, NUNC, Maxisorp, Roskilde, Denmark.i M-7504, Sigma-Aldrich, Steinheim, Germany.¶ Sanquin, Amsterdam, The Netherlands.# T-2885, Sigma-Aldrich.** Sunrise, TECAN, Salzburg, Austria.†† Cotinine Micro-Plate ELA Kit, STC Diagnostics, Bethlehem, PA.‡‡ Dade Behring, Amersfoort, The Netherlands.§§ SPSS package version 9.0, SPSS, Chicago, IL.

J Periodontol • November 2005 Maffei, Brouwer, Dolman, van der Velden, Roos, Loos

1883

coefficients for MBL levels in relation to cotinine levelsor cig/day were determined, controlling for periodon-tal status, age, gender, education, and ethnicity. Par-tial correlation coefficients for MBL in relation to CRPwere determined similarly, but also including smokingstatus. Possible differences in the prevalence of MBLdeficiency were tested with the chi-square test. For allanalyses, only P values <0.05 were considered statis-tically significant.

RESULTS

Patient characteristics are presented in Table 1. Themean ages for controls and moderate and severeperiodontitis and the distribution of individuals forgender and ethnicity did not differ among the threegroups. There were more non-smokers among thecontrols and more heavy smokers among the severeperiodontitis patients. The total number of teeth wascomparable among the three groups. Severe peri-odontitis patients had on average 13.7 teeth with

‡50% bone loss, while moderate periodontitis patientshad on average 2.4 teeth with ‡50% bone loss.

Among all subjects, plasma levels of MBL rangedbetween undetectable (<0.05 mg/ml) and 8.7 mg/ml.The mean values – standard deviation of MBL for thecontrol and moderate and severe periodontitis groupswere 1.6 – 1.9, 1.4 – 1.2, and 1.6 – 1.2 mg/ml, respec-tively. Medians and interquartile ranges of the MBLplasma levels among three groups are depicted inFigure 1A. Statistical analysis showed that the MBLplasma levels did not differ among the three groupsof subjects (P = 0.755); periodontitis patients did nothave elevated or decreased plasma levels of MBL incomparison to controls. From the statistical model, itappeared that smoking status was a highly significantcofactor (P = 0.002), level of education was a marginalsignificant cofactor (P = 0.04), and other covariateswere not significant. MBL was analyzed in relation toCRP; there was no correlation to plasma levels of CRP(r = 0.030 and P = 0.682).

The significance of smoking status in relation toMBL was unexpected. To investigate the relationshipbetween MBL and smoking, we analyzed the MBLplasma levels for the three smoking categories amongthe total study population. The mean MBL values fornon-, light, and heavy smokers were 1.4 – 1.5, 1.4 –1.4, and 2.2 – 1.6 mg/ml, respectively. A summaryof MBL values in the different smoking categories ispresented in Figure 1B. Since the overall ANCOVAmodel showed a significant difference (P = 0.002),we could analyze differences between any two groups:in post-hoc testing, it appeared that heavy smokers,irrespective of periodontal disease status, presentedwith higher levels of MBL than non-smokers. The factthat the MBL levels in smokers and non-smokersare independent of periodontitis can be seen in Figure1C. Irrespective of periodontal disease status, higherMBL in heavy smokers was observed. To further ex-plore the relationship between MBL and smoking,we analyzed plasma cotinine levels and cig/day inrelation to MBL plasma concentration (Table 2). Thepartial correlation coefficients for MBL and serum co-tinine and MBL and cig/day were weak but significant(Table 2).

In our laboratory, we consider subjects with MBLlevels <0.8 mg/ml as MBL deficient. With this defini-tion, we studied MBL deficiency in relation to peri-odontitis. In total, 89 (41%) of the study subjectswere MBL deficient. As shown in Table 3, the distribu-tion of MBL-deficient subjects was not statisticallydifferent among control and moderate and severeperiodontitis subjects. There was no increased preva-lence of MBL-deficient subjects in periodontitis.

In the next set of analyses, we excluded the MBL-deficient subjects and again investigated MBL levelsin the remaining study population (N = 130). The

Table 1.

Characteristics of the Study Population(mean – SD or numbers [%] of subjects)

Control

(N = 104)

Moderate

Periodontitis

(N = 70)

Severe

Periodontitis

(N = 45)

Age 40.7 – 12.0 43.1 – 11.1 42.0 – 8.0

Gender

Male 41(39%) 28 (40%) 20 (44%)

Female 63 (61%) 42 (60%) 25 (56%)

Ethnicity

Non-white 10 (10%) 11 (16%) 9 (20%)

White 94 (90%) 59 (84%) 36 (80%)

Education

<high school 21 (20%) 7 (10%) 9 (21%)

= high school 26 (25%) 32 (46%) 24 (55%)

>high school 57 (55%) 30 (44%) 11 (25%)

Smoking*

Non-smoker 80 (77%) 45 (64%) 21 (47%)

Light smoker 10 (10%) 9 (13%) 8 (18%)

Heavy smoker 14 (13%) 16 (23%) 16 (36%)

Number of teeth

Total 27.8 – 1.8 26.3 – 3.5 26.9 – 2.6

Without bone loss 27.8 – 1.8 11.0 – 7.4 2.7 – 3.3

With ‡one-thirdbone loss

0 5.2 – 4.1 16.6 – 4.9

With ‡50% bone loss 0 2.4 – 2 13.7 – 6

* P <0.001.

Mannan-Binding Lectin in Relation to Periodontitis and Smoking Volume 76 • Number 11

1884

mean plasma values for MBL were now 2.7 – 1.9, 2.0 –1.1, and 2.3 – 0.9 mg/ml for the control and moderateand severe periodontitis groups, respectively. Themedians and interquartile ranges for the three groupson the basis of periodontal disease are summarized inFigure 2A. The statistical analysis for non-deficientsubjects showed no increase or decrease in MBL lev-els in periodontitis. Again, in the latter analysis, smok-ing status was a significant cofactor (P = 0.011). Also,among the non-deficient subjects, MBL and CRP didnot correlate (r = 0.151 and P = 0.113). In Figure2B, we present MBL plasma levels for non-deficientsubjects subdivided into non-, light, and heavy smok-ers. Heavy smokers had again the highest MBLplasma levels. Also, in the MBL non-deficient sub-jects, the higher plasma levels of MBL were indepen-dent of periodontal disease status (Fig. 2C). Thepartial correlation coefficients among the non-deficient individuals for plasma cotinine levels andcig/day in relation to MBL levels were weak but signif-icant; the values were 0.197 and 0.236, respectively(Table 2).

Figure 1.Box and whisker plots for MBL plasma levels in the total studypopulation for (A) control subjects and moderate and severeperiodontitis patients; (B) non-, light, and heavy smokers; and(C) as in panel B but subgrouped for control and periodontitis. Thehorizontal line inside each box (interquartile range) indicates themedian (50th percentile). The P values for the overall ANCOVAs areindicated inside the graphs. *P = 0.003 in post-hoc test, afterBonferroni adjustment.

Table 2.

Partial Correlation Coefficients (r) andCorresponding P Values for the Relation ofMBL With the Smoking Parameters SerumCotinine and Self-Reported Cig/Day

r P

All subjects

MBL: cotinine 0.193 0.008

MBL: cig/day 0.194 0.008

MBL non-deficient subjects

MBL: cotinine 0.197 0.041

MBL: cig/day 0.236 0.014

Table 3.

Distribution of MBL Deficiency AmongDisease Categories (numbers [%]of subjects)

Control

(N = 104)

Moderate

Periodontitis

(N = 70)

Severe

Periodontitis

(N = 45)

Total

(N = 219)

MBLdeficient*

47 (45%) 26 (37%) 16 (36%) 89 (41%)

MBL non-deficient

57 (55%) 44 (63%) 29 (64%) 130 (59%)

The distribution of deficient subjects did not differ among disease categories(P = 0.216).* Plasma levels <0.8 mg/ml.

J Periodontol • November 2005 Maffei, Brouwer, Dolman, van der Velden, Roos, Loos

1885

Finally, we explored MBL levels in relation toimportant infectious agents in periodontitis patients.The levels of MBL, for all patients as well as for MBLnon-deficient patients, were not different whether ornot patients were infected with A. actinomycetemco-mitans, P. gingivalis, or both (data not shown). Table4 presents the distribution of MBL deficiency amongsubjects positive or negative for A. actinomycetemco-mitans, P. gingivalis, or the two bacteria together.These distributions of MBL deficiency were not statis-tically different in relation to the bacterial prevalences,indicating that infection with these periodontal patho-gens was not related to MBL status.

DISCUSSION

MBL has been suggested to be an important proteinof the innate immune system. It can act not onlyas an opsonin but also, importantly, can activatethe classical complement pathway in an antibody-independent manner (the MBL or lectin complementpathway). For example, MBL plays a role at the time ofprimary contact with a pathogen before an immuno-globulin M (IgM) antibody response can be mounted(ante-antibody hypothesis).15 Also, during the immu-nological window of vulnerability, between approxi-mately 6 and 18 months of age, MBL is consideredto play a key role.15 The interaction of MBL withMASPs is analogous to the interaction of the comple-ment factor C1q with complement factors C1rand C1s in the initial steps of the classical comple-ment activation pathway. In the classical complementpathway, C1q, C1r, and C1s, bound to immune-complexes, interact with C4 to activate the cascade.In a similar fashion, MBL complexed with MASPsand the carbohydrate patterns on bacteria interactswith C4.16

In periodontitis, in addition to infection with peri-odontal pathogens, host and environmental factorsplay a role in the susceptibility for and the severityof the disease. The present study is, to our knowledge,the first to investigate possible correlations betweenMBL and periodontitis. The results of our analyses re-vealed that MBL plasma levels did not differ betweenhealthy control subjects and periodontitis patients.Also, no relation with the severity of periodontitiswas found. These findings were the same, and some-what surprising, when MBL-deficient subjects wereexcluded as we expected MBL to behave as a (weak)acute-phase protein.36 The promoter region of theMBL gene contains several acute-phase responseelements.43 Previous studies in periodontitis patientshave shown modestly elevated levels of CRP, a proto-typical acute-phase protein,37,44 and the level ofplasma MBL was found to increase 1.5- to 3-fold inpatients undergoing major surgery and in patientswith malaria.34 However, among the current study

Figure 2.Box and whisker plots for MBL plasma levels in the non-deficientindividuals for (A) control subjects and moderate and severeperiodontitis patients; (B) non-, light, and heavy smokers; and(C) as in panel B but subgrouped for control and periodontitis. Thehorizontal line inside each box (interquartile range) indicates themedian (50th percentile). The P values for the overall ANCOVAs areindicated inside the graphs. *P = 0.003 in post-hoc test, afterBonferroni adjustment; †P = 0.009 in post-hoc test, after Bonferroniadjustment.

Mannan-Binding Lectin in Relation to Periodontitis and Smoking Volume 76 • Number 11

1886

subjects, no significant correlations were found be-tween CRP and MBL.

While plasma levels of MBL are on average 1.2 to1.6 mg/ml among healthy blood donors,27,28 it hasbeen suggested that subjects with levels below 0.5to 1.00 mg/ml are considered MBL deficient. However,the threshold of deficiency is still a matter of debate.28

Deficiency may be related to genetic variation. Sev-eral gene polymorphisms have been reported in theliterature: three mutations in exon one of the MBLgene (codons 52, 54, and 57) can cause structuralmodifications in the protein, resulting in low serumlevels. In addition, promoter polymorphisms cancause decreased production of the protein.15 MBLgene mutations and impaired serum levels have beenassociated with several infections in adults: individu-als who are homozygous for structural mutations ofthe MBL genotype have been reported to suffer fromrecurrent lung infections, recurrent otitis media, diar-rhea, and septicemia.29,30 However, more recentlyit was found that the frequency of MBL deficiency inpatients with signs of infections did not differ from thatof controls.31,32

In general, one-third of the white population is con-sidered MBL deficient.45 In our study population, theoverall frequency of MBL deficiency was 41%, i.e.,45% for healthy control subjects and 37% for peri-odontitis patients. The prevalence of MBL deficiencyin healthy subjects is somewhat higher than reportedin the literature. Further genetic studies are needed totype the current population and relate plasma levelswith genetic polymorphisms. Nevertheless, MBL defi-ciency at the functional protein level was not as-

sociated with (severe) periodontitis; therefore,according to our results, MBL deficiency may not pre-dispose individuals to develop periodontitis. MBL lev-els and deficiency were also not related to infectionwith the two most common periodontal pathogens,A. actinomycetemcomitans and P. gingivalis.

Increased MBL plasma levels in heavy smokers anda correlation of MBL with the number of cigarettes perday and cotinine levels were unexpected findings.Cigarette smoking has a strong impact on the generalcondition of individuals and has systemic effects. Forexample, in a survey including 4,516 men and womenaged 25 to 74 years, it was found that, specifically inmen, chronic smoking was associated with increasedwhite blood cell counts, elevated levels of fibrinogenand CRP, and increased plasma viscosity, comparedto never-smokers.46 Cigarette smoking is also as-sociated with a lack of increase in serum levels ofIgG in chronic infectious processes, including peri-odontitis.13,40,47-49 In our population, MBL plasmalevels were significantly increased in heavy smokers(Figs. 1B, 1C, 2B, and 2C). This observation hasnot been reported before in the literature. One expla-nation for this finding might be that heavy smokinghas enough impact to upregulate MBL production,even though MBL is considered a weak acute phaseprotein. Alternatively, it is interesting to speculate thatMBL levels in smokers are increased as a compensa-tory mechanism for the activation of the classicalcomplement pathway during (chronic) infections; incontrast to non-smokers, it has been established thatlevels of IgG in smokers do not increase during(chronic) infections.13,40,47-49

CONCLUSION

MBL plasma levels and MBL deficiency do not seem toaffect the susceptibility to and severity of periodonti-tis. Nevertheless, MBL plasma levels are increased insmokers, and smokers have more severe periodonti-tis. The underlying mechanisms for this correlationare not understood. The role of MBL and the activationof the MBL complement pathway during inflamma-tory processes such as periodontitis need to be eluci-dated in smokers and non-smokers.

REFERENCES1. Socransky SS, Haffajee AD, Cugini MA, Smith C, Kent

RL Jr. Microbial complexes in subgingival plaque.J Clin Periodontol 1998;25:134-144.

2. Haffajee AD, Socransky SS. Microbial etiologicalagents of destructive periodontal diseases. Periodontol2000 1994;5:78-111.

3. Brown LJ, Oliver RC, Loe H. Evaluating periodontalstatus of US employed adults. J Am Dent Assoc1990;121:226-232.

4. Brown LJ, Brunelle JA, Kingman A. Periodontal statusin the United States, 1988-1991: Prevalence, extent,

Table 4.

Distribution of MBL-Deficient Patients(plasma concentration <0.8 mg/ml;numbers [%] of subjects)

Culture

Negative

Culture

Positive

A. actinomycetemcomitans* N = 82 N = 33

MBL deficient 27 (33%) 15 (45%)

MBL non-deficient 55 (67%) 18 (54%)

P. gingivalis† N = 51 N = 64

MBL deficient 17 (33%) 25 (39%)

MBL non-deficient 34 (67%) 39 (61%)

A. actinomycetemcomitansand P. gingivalis‡

N = 40 N = 75

MBL deficient 12 (30%) 30 (40%)

MBL non-deficient 28 (70%) 45 (60%)

* P = 0.284.† P = 0.563.‡ P = 0.316.

J Periodontol • November 2005 Maffei, Brouwer, Dolman, van der Velden, Roos, Loos

1887

and demographic variation. J Dent Res 1996;75 SpecNo:672–683.

5. Gjermo P. Epidemiology of periodontal disease inEurope. J Parodontol d’Implantol Orale 1998;17:111-121.

6. Page RC, Offenbacher S, Schroeder HE, Seymour GJ,Kornman KS. Advances in the pathogenesis of peri-odontitis: Summary of developments, clinical implica-tions and future directions. Periodontol 2000 1997;14:216-248.

7. Loos BG, van der Velden U. Genetics in relation toperiodontitis. In: Lindhe J, Karring T, Lang NP, eds.Clinical Periodontology and Implant Dentistry. Oxford:Blackwell Munksgaard; 2003:387-399.

8. Bergstrom J. Cigarette smoking as risk factor inchronic periodontal disease. Community Dent OralEpidemiol 1989;17:245-247.

9. Haber J, Wattles J, Crowley M, et al. Evidence forcigarette smoking as a major risk factor for periodon-titis. J Periodontol 1993;64:16-23.

10. Bouclin R, Landry RG, Noreau G. The effects of smok-ing on periodontal structures: A literature review.J Can Dent Assoc 1997;63:356, 360-363.

11. Xu L, Loos BG, Craandijk J, et al. Teeth with peri-odontal bone loss, cigarette smoking and plasmacotinine levels. J Int Acad Periodontol 2002;4:39-43.

12. MacFarlane GD, Herzberg MC, Wolff LF, Hardie NA.Refractory periodontitis associated with abnormalpolymorphonuclear leukocyte phagocytosis and cig-arette smoking. J Periodontol 1992;63:908-913.

13. Gunsolley JC, Pandey JP, Quinn SM, Tew J, SchenkeinHA. The effect of race, smoking and immunoglobulinallotypes on IgG subclass concentrations. J PeriodontalRes 1997;32:381-387.

14. Loos BG, Roos MT, Schellekens PT, van der Velden U,Miedema F. Lymphocyte numbers and function inrelation to periodontitis and smoking. J Periodontol2004;75:557-564.

15. Turner MW, Hamvas RM. Mannose-binding lectin:Structure, function, genetics and disease associations.Rev Immunogenet 2000;2:305-322.

16. Janeway CA, Travers P, Walport M, Shlomchik M.Immunobiology, 5th ed. New York: Garland Publish-ing; 2001.

17. Jack DL, Dodds AW, Anwar N, et al. Activation ofcomplement by mannose-binding lectin on isogenicmutants of Neisseria meningitidis serogroup B. J Im-munol 1998;160:1346-1353.

18. Neth O, Jack DL, Dodds AW, et al. Mannose-bindinglectin binds to a range of clinically relevant micro-organisms and promotes complement deposition.Infect Immun 2000;68:688-693.

19. Townsend R, Read RC, Turner MW, Klein NJ, Jack DL.Differential recognition of obligate anaerobic bacteriaby human mannose-binding lectin. Clin Exp Immunol2001;124:223-228.

20. Bramanti TE, Wong GG, Weintraub ST, Holt SC.Chemical characterization and biologic properties oflipopolysaccharide from Bacteroides gingivalis strainsW50, W83, and ATCC 33277. Oral Microbiol Immunol1989;4:183-192.

21. Califano JV, Schenkein HA, Tew JG. Immunodomi-nant antigens of Actinobacillus actinomycetemcomi-tans serotypes a and c in high-responder patients. OralMicrobiol Immunol 1991;6:228-235.

22. Farquharson SI, Germaine GR, Gray GR. Isolation andcharacterization of the cell-surface polysaccharides of

Porphyromonas gingivalis ATCC 53978. Oral Micro-biol Immunol 2000;15:151-157.

23. Schifferle RE, Reddy MS, Zambon JJ, Genco RJ,Levine MJ. Characterization of a polysaccharide anti-gen from Bacteroides gingivalis. J Immunol 1989;143:3035-3042.

24. Zambon JJ, Slots J, Miyasaki K, et al. Purification andcharacterization of the serotype c antigen from Actino-bacillus actinomycetemcomitans. Infect Immun 1984;44:22-27.

25. Jack DL, Klein NJ, Turner MW. Mannose-bindinglectin: Targeting the microbial world for complementattack and opsonophagocytosis. Immunol Rev 2001;180:86-99.

26. Kuhlman M, Joiner K, Ezekowitz RA. The humanmannose-binding protein functions as an opsonin.J Exp Med 1989;169:1733-1745.

27. Garred P, Thiel S, Madsen HO, et al. Gene frequencyand partial protein characterization of an allelic variantof mannan binding protein associated with low serumconcentrations. Clin Exp Immunol 1992;90:517-521.

28. Crosdale DJ, Ollier WE, Thomson W, et al. Mannosebinding lectin (MBL) genotype distributions with re-lation to serum levels in UK Caucasoids. Eur J Im-munogenet 2000;27:111-117.

29. Garred P, Madsen HO, Hofmann B, Svejgaard A. In-creased frequency of homozygosity of abnormal mannan-binding-protein alleles in patients with suspectedimmunodeficiency. Lancet 1995;346:941-943.

30. Summerfield JA, Sumiya M, Levin M, Turner MW.Association of mutations in mannose binding proteingene with childhood infection in consecutive hospitalseries. BMJ 1997;314:1229-1232.

31. Cedzynski M, Szemraj J, Swierzko AS, et al. Mannan-binding lectin insufficiency in children with recurrentinfections of the respiratory system. Clin Exp Immunol2004;136:304-311.

32. Aittoniemi J, Rintala E, Miettinen A, Soppi E. Serummannan-binding lectin (MBL) in patients with infec-tions: Clinical and laboratory correlates. APMIS 1997;105:617-622.

33. Dahl M, Tybjaerg-Hansen A, Schnohr P, NordestgaardBG. A population-based study of morbidity and mor-tality in mannose-binding lectin deficiency. J Exp Med2004;199:1391-1399.

34. Thiel S, Holmskov U, Hviid L, Laursen SB, JenseniusJC. The concentration of the C-type lectin, mannan-binding protein, in human plasma increases during anacute phase response. Clin Exp Immunol 1992;90:31-35.

35. Taylor ME, Brickell PM, Craig RK, Summerfield JA.Structure and evolutionary origin of the gene encodinga human serum mannose-binding protein. Biochem J1989;262:763-771.

36. Petersen SV, Thiel S, Jensenius JC. The mannan-binding lectin pathway of complement activation: Biol-ogy and disease association. Mol Immunol 2001;38:133-149.

37. Loos BG, Craandijk J, Hoek FJ, Wertheim-van DillenPME, van der Velden U. Elevation of systemic markersrelated to cardiovascular diseases in the peripheralblood of periodontitis patients. J Periodontol 2000;71:1528-1534.

38. Hutter JW, van der Velden U, Varoufaki A, et al. Lowernumbers of erythrocytes and lower levels of hemo-globin in periodontitis patients compared to controlsubjects. J Clin Periodontol 2001;28:930-936.

Mannan-Binding Lectin in Relation to Periodontitis and Smoking Volume 76 • Number 11

1888

39. van Winkelhoff AJ, Loos BG, van der Reijden WA,van der Velden U. Porphyromonas gingivalis,Bacteroides forsythus and other putative periodontalpathogens in subjects with and without periodon-tal destruction. J Clin Periodontol 2002;29:1023-1028.

40. Graswinckel JEM, van der Velden U, van WinkelhoffAJ, Hoek FJ, Loos BG. Plasma antibody levels inperiodontitis patients and controls. J Clin Periodontol2004;31:562-568.

41. van Winkelhoff AJ, Carlee AW, de Graaff J. Bacter-oides endodontalis and other black-pigmented Bac-teroides species in odontogenic abscesses. InfectImmun 1985;49:494-497.

42. Mendall MA, Patel P, Asante M. Relation of serumcytokine concentrations to cardiovascular risk fac-tors and coronary heart disease. Heart 1997;78:273-277.

43. Ezekowitz RA, Day LE, Herman GA. A human man-nose-binding protein is an acute phase reactant thatshares sequence homology with other vertebratelectins. J Exp Med 1988;167:1034-1046.

44. Fredriksson M, Bergstrom K, Asman B. IL-8 andTNF-a from periferal neutrophils and acute-phaseproteins in periodontitis. J Clin Periodontol 2002;29:123-128.

45. Steffensen R, Thiel S, Varming K, Jersild C, JenseniusJC. Detection of structural gene mutations and pro-moter polymorphism in the manna-binding lectin(MBL) gene by polymerase chain reaction with se-quence-specific primers. J Immunol Methods 2000;241:33-42.

46. Frohlich M, Sund M, Lowel H, et al. Independentassociation of various smoking characteristics withmarkers of systemic inflammation in men. Eur Heart J2003;24:1365-1372.

47. Gulsvik A, Fagerhoi MK. Smoking and immunoglob-ulin levels. Lancet 1979;1:449.

48. Quinn SM, Zhang JB, Gunsolley JC, Schenkein HA,Tew JG. The influence of smoking and race on adultperiodontitis and serum IgG2 levels. J Periodontol1998;69:171-177.

49. Moszczynski P, Zabinski Z, Moszczynski P Jr, et al.Immunological findings in cigarette smokers. ToxicolLett 2001;118:121-127.

Correspondence: Dr. Bruno Loos, Department of Peri-odontology, Academic Center for Dentistry Amsterdam(ACTA), Louwesweg 1, 1066 EA Amsterdam, The Nether-lands. Fax: 31-20-5188512; e-mail: b.g.loos@acta.nl.

Accepted for publication March 25, 2005.

J Periodontol • November 2005 Maffei, Brouwer, Dolman, van der Velden, Roos, Loos

1889