Short communication

NADPH-diaphorase activity in nerves and Schwanncells in the periodontal ligament of rat incisor teeth

Hiroyuki Ichikawa *, Tomosada Sugimoto

Second Department of Oral Anatomy, Okayama University Dental School, Okayama, Japan

Abstract

The lingual portion of the incisor periodontal ligament demonstrated activity for nicotinamide adenosine dinucleo-tide phosphate (NADPH)-diaphorase. Schwann cells surrounding Ru�ni-like endings coexpressed NADPH-diaphor-

ase activity and immunoreactivity for inducible nitric oxide synthase. NADPH-diaphorase-positive nerve ®breswhich coexpressed immunoreactivity for neuronal nitric oxide synthase were in contact with Schwann cells surround-ing Ru�ni-like endings or terminated as free nerve endings. Neural NADPH-diaphorase activity could not be found

in the tissues covering the labial portion of incisor tooth root. It is possible that nitric oxide in Schwann cells andnerves has functions speci®c to the incisor periodontal ligament. # 1998 Elsevier Science Ltd. All rights reserved

Key words: Nicotinamide adenosine dinucleotide phosphate-diaphorase, Periodontal ligament, Nitric oxide synthase, Ru�ni-like

ending, Free nerve ending, Incisor, Rat

NADPH-diaphorase, a histochemical marker for nitric

oxide synthase, occurs in the central and peripheral

nervous systems (Aimi et al., 1991; Alm et al., 1995;

Hisa et al., 1996; Kerezoudis et al., 1993b; Morris et

al., 1993; Paakkari and Lindsberg, 1995; Vincent and

Kimura, 1992). In the trigeminal ganglion, it is loca-

lized in small to medium-sized primary sensory neur-

ones (Aimi et al., 1991; Alm et al., 1995; Kerezoudis et

al., 1993b; Morris et al., 1993). These neurones also

coexpress immunoreactivities for substance P and

CGRP (Aimi et al., 1991). Because these two are con-

sidered to be putative transmitters for nociception in

primary sensory neurones, NADPH-diaphorase-posi-

tive trigeminal neurones may include primary nocicep-

tors. Recently, it has been demonstrated that

NADPH-diaphorase in neurones is neuronal nitric

oxide synthase, one of the enzyme's isoforms (Hisa et

al., 1996; Paakkari and Lindsberg, 1995). In oral tis-

sues, inhibition of this neuronal isoform reportedly

had e�ects on basal blood ¯ow and antidromic vasodi-

lation but not neurogenic plasma extravasation

(Kerezoudis et al., 1993a; 1994). On the other hand,

glial cells in the brainstem contain macrophage-type

inducible nitric oxide synthase and NADPH-diaphor-

ase (Galea et al., 1992; Paakkari and Lindsberg, 1995).

Nitric oxide synthesized by the inducible synthase has

been suggested to have various functions associated

with host defence and plasticity in glial cells, and

immunoregulatory roles of Schwann cells (Gold et al.,

1996; Paakkari and Lindsberg, 1995).

The periodontal ligament receives innervation from

primary sensory neurones. Their receptors include free

and Ru�ni-like endings (Anderson et al., 1970; Byers,

1985; Byers and Dong, 1989; Byers et al., 1986; Sato et

al., 1988; Silverman and Kruger, 1987; Wakisaka et

al., 1985). Periodontal free nerve endings display sub-

stance P and CGRP immunoreactivities, and are

thought to be nociceptors probably derived from small

neurones in the trigeminal ganglion (Anderson et al.,

1970; Byers, 1985; Byers and Dong, 1989; Silverman

and Kruger, 1987; Wakisaka et al., 1985). Neurones

with periodontal Ru�ni-like endings have their cell

bodies in the trigeminal mesencephalic nucleus or tri-

geminal ganglion (Byers, 1985; Byers and Dong, 1989).

The receptors on the mesencephalic nucleus are

Archives of Oral Biology 43 (1998) 167±171

0003-9969/98/$19.00 # 1998 Elsevier Science Ltd. All rights reserved

PII: S0003-9969(97 )00099-X

ARCHIVESOFORALBIOLOGY

* To whom all correspondence should be addressed.

Abbreviations: CGRP, calcitonin gene-related peptide,

NADPH, nicotinamide adenosine dinucleotide phosphate.

thought to be involved in monitoring tooth movement

and in the re¯ex control of mandibular movementsduring mastication (Byers et al., 1986). Ru�ni-likeendings of trigeminal neurones are mechanoreceptors

that are thought to be activated by touch, pressure andmovement of teeth during chewing, swallowing andspeech. Axon terminals in Ru�ni-like endings are cov-

ered with lamellar Schwann cells that are immuno-reactive for glia-speci®c S100 protein (Sato et al.,

1988). A previous study demonstrated that in the peri-odontal ligament of rat molar teeth these receptorswere devoid of NADPH-diaphorase activity

(Kerezoudis et al., 1993b). The structures and func-tions of periodontal ligaments are di�erent in ratmolar and incisor teeth, and NADPH-diaphorase ac-

tivity has never been reported in the incisor periodon-tal ligament. We have now examined NADPH-

diaphorase activity in the periodontal ligaments of ratincisor teeth in order to determine whether periodontalreceptors utilize nitric oxide. The coexpression of

NADPH-diaphorase activity with immunoreactivitiesfor neuronal and inducible isoforms of nitric oxidesynthase and S100, and the ultrastructure of inducible

nitric oxide synthase-immunoreactive components werealso investigated to characterize NADPH-diaphorase-

positive pro®les.Eight adult male Sprague±Dawley rats (200±300 g)

were used. Animals were anaesthetized with ether to

the level at which respiration was markedly suppressed,and transvascularly perfused with 50 ml of isotonic

saline (154 mM NaCl) followed by 500 ml of 4% for-maldehyde in 0.1 M phosphate bu�er (pH 7.4). Forthree animals to be examined by electron microscopy,

0.05% glutaraldehyde was added to the ®xative.Mandibles containing incisor teeth were dissected andmineralized with 4.13% EDTA disodium salt in 0.1 M

phosphate bu�er (pH 7.4) for 1 week at 48C. The tis-sues were soaked overnight in a phosphate-bu�ered

saline containing 20% sucrose, frozen-sectioned at12 mm, and mounted on gelatin-coated glass slides.For NADPH-diaphorase histochemistry, sagittal sec-

tions were incubated with 0.1 M phosphate bu�er (pH7.4) containing 0.1 mg/ml nitroblue tetrazolium(Sigma, U.S.A.) and 1.0 mg/ml b-NADPH (Sigma) for

2 hr at 378C.For the coexpression study, a double-immuno¯uor-

escence method was used. Sections were incubatedwith a mixture of rabbit anti-neuronal nitric-oxidesynthase serum (1:1000; Chemicon International, Inc.,

U.S.A.) and mouse monoclonal anti-S100 antibody(1:1000; Sigma) or with a mixture of rabbit anti-induci-

ble nitric-oxide synthase serum (1:1000; Santa CruzBiotechnology Inc., U.S.A.) and mouse monoclonalanti-S100 antibody, followed by incubation with a mix-

ture of lissamine rhodamine B chloride-conjugateddonkey antirabbit IgG (1:400; Jackson

ImmunoResearch Labs, U.S.A.) and ¯uorescein iso-

thiocyanate-conjugated donkey antimouse IgG (1:100;Jackson ImmunoResearch). Subsequent to photomi-croscopy of immuno¯uorescent pro®les, the coverslips

were removed, and the sections were stained forNADPH-diaphorase acitivity.

For electron microscopy, unfrozen 50 mm-thicksagittal sections were cut with a Microslicer (DosakaEM, Japan) and stained for inducible nitric-oxide

synthase immunoreactivity with an avidin±biotin±hor-seradish peroxidase complex method. The sectionswere incubated with the primary antibody (1:20000)

for 5 days at 48C followed by biotinylated goat anti-rabbit IgG and the avidin complex (Vector

Laboratories, U.S.A.). Following diaminobenzidinereaction, these sections were post®xed in 1% osmiumtetroxide in 0.1 M phosphate bu�er (pH 7.4), dehy-

drated through a graded series of alcohols, andembedded in Polybed 812. Ultrathin sections wereexamined after staining with lead citrate for 1 min.

In control experiments, the primary antibodies werepreabsorbed with appropriate proteins (50 mg/ml;

Santa Cruz Biotechnology for neuronal and induciblesynthases, Sigma for S100). No staining was observedin the controls. To examine whether the NADPH-dia-

phorase-positive ®bres observed were neural elements,the right inferior alveolar nerve was transected in onerat, and the mandible was stained for NADPH-dia-

phorase. At 7 days after the transection, virtually allpositively stained ®bres had disappeared from the peri-

odontal ligament of the mandibular teeth on the ipsi-lateral side. Thus, we consider that the NADPH-diaphorase-positive ®bres were nerve ®bres, and the

term ``nerve ®bres'' is used throughout this paper.NADPH-diaphorase activity was observed in cells

and nerve ®bres in the lingual portion of incisor peri-odontal ligament (Fig. 1A±C). These cells and ®breswere adjacent to the alveolar bone. The cells had var-

ious shapes, including round, oval and triangular, withor without processes (diameters of cells, 3±10 mm), andwere aggregated in close apposition to nerve bundles

and blood vessels (mean number2S.E.M of positivecells/section = 126217, n= 4). The reaction products

for this enzyme were restricted to the cytoplasm.NADPH-diaphorase-positive ®bres were abundant innerve bundles and had a ®ne, varicose appearance.

Some positive nerve ®bres were also observed accom-panying blood vessels. These ®bres left nerve bundlesand blood vessels, and surrounded NADPH-diaphor-

ase-positive cells (Fig. 1B) or terminated as free nerveendings (Fig. 1C). Nearly 30% (137/505) of NADPH-

diaphorase-positive cells were seen in close contactwith the positive nerve ®bres.The double-immuno¯uorescence method in combi-

nation with NADPH-diaphorase histochemistryrevealed coexpression of NADPH-diaphorase activity

H. Ichikawa, T. Sugimoto / Archives of Oral Biology 43 (1998) 167±171168

Fig. 1. Photomicrographs of NADPH-diaphorase activity (A±D, G), inducible nitric oxide synthase (iNOS) immunoreactivity (E, I,

J), S100-immunoreactivity (F) and neuronal nitric oxide synthase (nNOS) immunoreactivity (H) in the lingual portion of the incisor

periodontal ligament. NADPH-diaphorase-positive cells have various shapes with or without processes (arrows in A, B). NADPH-

diaphorase-positive nerve ®bres surround the positive cells (arrowheads in B) in close apposition to a blood vessel (bv in B) or ter-

minate as free nerve endings (arrowheads in C). (D±F) and (G, H) are the same ®elds of views. NADPH-diaphorase-positive cells

(arrows in D) coexpress both iNOS (arrows in E) and S100 immunoreactivities (arrows in F). The distribution of NADPH-diaphor-

ase-positive nerve ®bres (arrowheads in G) is very similar to that of nNOS-immunoreactive nerve ®bres (arrowheads in H). Arrows

in (G) and (H) indicate NADPH-diaphorase-positive cells (G) which are devoid of nNOS immunoreactivity (H). (I) and (J) show

electron micrographs of a Ru�ni-like ending obtained from adjacent sections. A terminal Schwann cell surrounding a Ru�ni-like

ending (s in I) contains dense reaction products for iNOS immunoreactivity (large arrowheads in I) and is covered with multiple

layers of basal lamina (small arrows in I). Large arrows in (I) and (J) indicate the same short projection of the axoplasm. Clusters

of small vesicles are seen at the base of the projection (small arrowheads in I and J). Scale bars: 50 mm (A±C), 100 mm (D±H) and

1 mm (I, J).

H. Ichikawa, T. Sugimoto / Archives of Oral Biology 43 (1998) 167±171 169

with immunoreactivities for inducible and neuronal

nitric oxide synthase and S100 in the periodontal liga-ment. All NADPH-diaphorase-positive cells coex-pressed the inducible synthase and S100

immunoreactivities. In addition, all cells immuno-reactive for inducible nitric oxide synthase coexpressedNADPH-diaphorase activity and S100 immunoreactiv-

ity (Fig. 1D±F). In nerve bundles, however, S100-im-munoreactive Schwann cells were devoid of NADPH-

diaphorase activity and inducible synthase immunor-eactivity. NADPH-diaphorase-positive cells werealways devoid of immunoreactivity for neuronal nitric

oxide synthase. The distributions of NADPH-diaphor-ase-positive nerve ®bres and neuronal nitric oxidesynthase-immunoreactive ones were identical (Fig. 1,

G, H). No nerve ®bres immunoreactive for the induci-ble synthase were observed in the periodontal liga-

ment.The coexpression study demonstrated that NADPH-

diaphorase-positive cells and cells immunoreactive for

inducible nitric oxide synthase had an identical distri-bution in the periodontal ligament of incisor teeth.

Thus, inducible synthase-immunoreactive cells wereexamined by an immunoelectron-microscopic methodto characterize NADPH-diaphorase-positive cells. The

clusters of inducible synthase-immunoreactive cells inthe periodontal ligament turned out to be Ru�ni-likeendings (Fig. 1, I). Terminal Schwann cells surround-

ing Ru�ni-like endings contained numerous pinocyto-tic vesicles and some mitochondria, and formed

lamellar sheets around the axoplasm. These Schwanncells were covered with multiple layers of basal lamina.The axoplasm was enriched with mitochondria (Fig. 1,

I) and formed short projections that extended into thesurrounding intercellular space (Fig. 1, J). Clusters ofsmall vesicles were often seen at the base of the projec-

tions. It was the Schwann cell and its process thatexpressed immunoreactivity for inducible nitric oxide

synthase in the Ru�ni-like endings, while the axo-plasm was devoid of it. At a higher magni®cation, theimmunoreaction products were distributed over mem-

branes of pinocytotic vesicles and the cytoplasm inSchwann cells.Neural NADPH-diaphorase activity could not be

observed in the tissues covering the labial portion ofthe roots of incisor teeth.

We demonstrate that the periodontal ligament of ratincisor teeth contains neural NADPH-diaphorase ac-tivity. NADPH-diaphorase-positive cells that coex-

pressed immunoreactivities for inducible nitric oxidesynthase and S100 were distributed in the lingual por-

tion of the ligament. Our electron-microscopic analysisfor the inducible synthase immunoreactivity indicatedthat these NADPH-diaphorase-positive cells were iden-

tical to Schwann cells and their processes associatedwith Ru�ni-like endings. This is supported by previous

®ndings that glial cells in the brainstem contained

NADPH-diaphorase activity and immunoreactivity forinducible nitric oxide synthase (Galea et al., 1992;Paakkari and Lindsberg, 1995) and that Schwann cells

in periodontal Ru�ni-like endings exhibited S100immunoreactivity (Sato et al., 1988). Because Schwanncells in nerve bundles were devoid of NADPH-dia-

phorase activity and the inducible synthase immunor-eactivity, these cells are probably unable to synthesize

nitric oxide. Thus, nitric oxide may be associated withthe functions of Schwann cells speci®c to Ru�ni-likeendings surrounding the incisor periodontal ligament.

We also demonstrate that nerve ®bres in the incisorperiodontal ligament coexpress NADPH-diaphorase

activity and immunoreactivity for neuronal nitric oxidesynthase. Because both primary sensory neurones inthe trigeminal ganglion and parasympathetic post-

ganglionic neurones in the otic and pterygopalatineganglia contain NADPH-diaphorase activity (Aimi etal., 1991; Alm et al., 1995; Morris et al., 1993;

Kerezoudis et al., 1993b), the origin of these positivelystained periodontal nerve ®bres is still unclear.

However, their sensory nature is probably suggestedby the distribution of their terminals; NADPH-dia-phorase-positive nerve ®bres terminated as free nerve

endings or were in contact with Schwann cells inRu�ni-like endings. This may be supported by our

control ®ndings that virtually all NADPH-diaphorase-positive nerve ®bres in the incisor periodontal ligamentdisappeared after transection of the inferior alveolar

nerve. The possibility that these nerve ®bres originatefrom the trigeminal mesencephalic nucleus is excluded,because primary sensory neurones in that nucleus are

devoid of the enzyme activity (Vincent and Kimura,1992) and because periodontal receptors from that

nucleus other than Ru�ni-like endings have not beenreported (Byers et al., 1986). Thus, it can be deducedthat NADPH-diaphorase-positive nerve ®bres in the

periodontal ligament at least partly originate from thetrigeminal ganglion.

The coexpression of CGRP and substance P im-munoreactivities by the NADPH-diaphorase-positivetrigeminal cells (Aimi et al., 1991) and the demon-

strated free nerve endings strongly suggest that at leastsome of the NADPH-diaphorase-positive trigeminalneurones are involved in nociception. It is possible that

nitric oxide synthesized by trigeminal neurones isinvolved in sensory signal transduction (Paakkari and

Lindsberg, 1995). On the other hand, the preferentialdistribution of NADPH-diaphorase activity and nitricoxide synthase immunoreactivity in the periodontal

ligament of the incisor but not molar teeth maysuggest that the neural nitric oxide is related to thespeci®c tissue environment. Because the adult rodent

incisor is continually erupting, the sensory endings inits periodontal ligament have to accommodate more

H. Ichikawa, T. Sugimoto / Archives of Oral Biology 43 (1998) 167±171170

tissue reorganization than is the case for molars. Nitricoxide may thus be involved in the plasticity of sensory

nerve endings in the rapidly erupting incisor periodon-tal ligament.

References

Aimi, Y., Fujimura, M., Vincent, S. R., Kimura, H. (1991)

Localization of NADPH-diaphorase-containing neurons in

sensory ganglia of the rat. J. Comp. Neurol. 30, 382±392.

Alm, P., Uvelius, B., Ekstrom, J., Holmqvist, B., Larsson, B.,

Andersson, K-E. (1995) Nitric oxide synthase-containing

neurons in rat parasympathetic, sympathetic and sensory

ganglia: a comparative study. Histochem. J. 27, 819±831.

Anderson, D. J., Hannam, A. G., Matthews, B. (1970)

Sensory mechanisms in mammalian teeth and their support-

ing structures. Physiol. Rev 50, 171±195.

Byers, M. R. (1985) Sensory innervation of periodontal liga-

ment of rat molars consists of uncapsulated Ru�ni-like

mechanoreceptors and free nerve endings. J. Comp. Neurol.

231, 500±518.

Byers, M. R., Dong, W. K. (1989) Comparison of trigeminal

receptor location and structure in the periodontal ligament

of di�erent types of teeth from the rat, cat and monkey. J.

Comp. Neurol. 279, 117±127.

Byers, M. R., O'Connor, T. A., Martin, R. F., Dong, W. K.

(1986) Mesencepharic trigeminal sensory neurons of cat;

Axon pathways and structure of mechanoreceptive endings

in periodontal ligament. J. Comp. Neurol. 250, 181±191.

Galea, E., Feinstein, D. L., Reis, D. J. (1992) Induction of

calcium-independent nitric oxide synthase activity in pri-

mary rat glial cultures. Proc. Natl. Acad. Sci. U.S.A. 89,

10945±10949.

Gold, R., Zielasek, J., Kiefer, R., Toyka, K. V., Hartung, H.

P. (1996) Secretion of nitrite by Schwann cells and its e�ect

on T-cell activation. Cell. Immunol. 168, 69±77.

Hisa, Y., Tadaki, N., Uno, T., Koike, S., Tanaka, M.,

Okamura, H., Ibata, Y. (1996) Nitrergic innervation of the

rat larynx measured by nitric oxide synthase immunohisto-

chemistry and NADPH-diaphorase histochemistry. Ann.

Otol. Rhinol. Laryngol. 105, 550±554.

Kerezoudis, N. P., Olgart, L., Edwall, L. (1993a) Di�erential

e�ects of nitric oxide synthasis inhibition on basal blood

¯ow and antidromic vasodilation in rat oral tissues. Eur. J.

Pharmacol. 241, 209±219.

Kerezoudis, N. P., Olgart, L., Edwall, L. (1994) Involvement

of substance P but not nitric oxide or calcitonin gene-re-

lated peptide in neurogenic plasm extravasation in rat inci-

sor pulp and lip. Arch. Oral. Biol. 39, 769±774.

Kerezoudis, N. P., Olgart, L., Fried, K. (1993b) Localization

of NADPH-diaphorase activity in the dental pulp, period-

ontium and alveolar bone of the rat. Histochemistry 100,

319±322.

Morris, R., Southam, E., Gittins, S. R., Garthwaite, J. (1993)

NADPH-diaphorase staining in autonomic and somatic

cranial ganglia of the rat. NeuroReport 4, 65±68.

Paakkari, I., Lindsberg, P. (1995) Nitric oxide in the central

nervous system. Ann. Med. 27, 369±377.

Sato, O., Maeda, T., Kobayashi, S., Iwanaga, T., Fujita, T.,

Takahashi, Y. (1988) Innervation of periodontal ligament

and dental pulp in rat incisors. An immunohistochemical

investigation of neuro®lament protein and glia speci®c S100

protein. Cell Tiss. Res. 251, 13±21.

Silverman, J. D., Kruger, L. (1987) An interpretation of den-

tal innervation based upon the pattern of calcitonin gene-

related peptide (CGRP) immunohistochemistry. Somatosen.

Res. 5, 157±175.

Vincent, S. R., Kimura, H. (1992) Histochemical mapping of

nitric oxide synthase in the rat brain. Neuroscience 46, 755±

784.

Wakisaka, S., Nishikawa, S., Ichikawa, H., Matsuo, S.,

Takano, Y., Akai, M. (1985) The distribution and origin of

substance P-like immunoreactivity in the rat molar pulp

and periodontal tissues. Arch. oral. Biol. 30, 813±818.

H. Ichikawa, T. Sugimoto / Archives of Oral Biology 43 (1998) 167±171 171