Analysis of Protein Immunoexpression and Its Interrelationship in the Pathogenesis of Odontomas and Ameloblastic Fibro-Odontomas: A Systematic Review

Abstract

Odontomas and ameloblastic fibro-odontomas (AFOs) are the result of a developmental anomaly of odontogenic tissues. A literature review of proteins immunoexpressed in odontomas and AFOs was conducted in order to determine which proteins are involved in the pathogenesis of these lesions. AFO was changed to early odontoma in the 2017 WHO classification and will also be discussed in this article. A literature search was performed in the following electronic databases: PubMed/MEDLINE, Web of Science, Scopus, EMBASE, Lilacs, Cochrane Collaboration Library, and Science Direct. The research question was developed according to the population, intervention, comparison, and outcome (PICO) framework: Which proteins are related to the differentiation of odontomas and what is their interrelationship with AFOs? Thirty articles met all inclusion criteria and were selected for this systematic review, totaling 355 cases of odontomas and 43 cases of AFO. Similar immunoexpression was observed in odontomas and AFOs. Immunoexpression of proteins involved in cell differentiation was higher in compound odontomas than in complex odontomas. Proteins involved in histodifferentiation and enamel formation were more frequent in odontomas. The immunoexpression of enamel matrix proteins differs between odontomas and tooth germs, with their persistence being related to the development of odontomas. Compound odontomas exhibit the highest immunoexpression of proteins involved in cellular histodifferentiation and the Wnt/beta-catenin pathway is involved in tumor formation.

Electronic supplementary material

The online version of this article (10.1007/s12105-020-01260-x) contains supplementary material, which is available to authorized users.

Introduction

Odontomas are benign mixed ectomesenchymal tumors characterized by hamartomatous growth of dental tissues in variable proportions [1]. These tumors are commonly found in the permanent dentition of children and young adults and are rarely associated with first molars [2, 3]. In most cases, odontomas are intraosseous lesions that can inhibit the eruption of the permanent tooth. Complete development within soft tissues occurs in a few cases and these lesions are called gingival or peripheral odontomas [4].

The etiology of odontomas is unknown but they are possibly the result of genetic mutations in the tooth germ. Lesions formerly referred to as ameloblastic fibro-odontomas (AFOs) are currently considered immature stages of an odontoma [5]. AFO was changed to early odontoma in the 2017 WHO classification and will also be discussed in this article. Some authors suggest that a supernumerary tooth is the result of different expression of the same pathological process that gives origin to an odontoma. However, although this idea may be applicable in some situations, supernumerary teeth do not cause a slow-growing mass as observed in odontomas. The latter are considered benign tumors arising from odontogenic tissue that have an indolent behavior [6].

In view of the above considerations, the aim of this study was to review for the first time the current literature regarding studies that investigate the immunoexpression of proteins and their interrelationship in the pathogenesis of odontomas and AFOs.

Material and Methods

The present literature review followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement [7].

Search Strategies

A literature search was performed from 7 to 14 August 2020 in the following electronic databases: PubMed/MEDLINE, Web of Science, Scopus, EMBASE, Lilacs, Cochrane Collaboration Library, and Science Direct. The following research question was developed according to the population, intervention, comparison, and outcome (PICO) framework: “Which proteins are related to the differentiation of odontomas and what is their interrelationship with ameloblastic fibro-odontomas? A systematic review was conducted. The study was previously registered in the 2020 International Prospective Register of Systematic Reviews (PROSPERO) under protocol No. CRD42020192372.

The search strategy was based on combinations of the following keywords: (Odontoma [mesh] OR Fibro-Odontoma [tw] OR Compound Odontoma [tw] OR Complex Odontoma [tw] OR Ameloblastic Fibro-odontoma [tw] OR Tooth [mesh] OR Teeth [tw] or Odontogenesis [mesh] OR Odontogeneses [tw]) AND (Immunohistochemistry [mesh] OR Immunolabeling Technique [tw] OR Immunolabeling Technic [tw] OR Immunogold Technique [tw] OR Immunogold Technic [tw] OR Immunohistocytochemistry [tw] OR Immunogold-Silver Technique [mesh] OR Immunogold-Silver Technic [tw] OR Immunocytochemistry [tw]) AND (Odontogenic Tumor [mesh] OR Tumor, Odontogenic [tw] OR Dental Tissue Neoplasm [tw] OR Neoplasm, Dental Tissue [tw] OR Tissue Neoplasm, Dental [tw]). The search was carried out without time and language restrictions.

Study Selection

The titles and abstracts of all articles identified by the electronic searches were read independently by the authors. In a second round, the full text of studies apparently meeting the inclusion criteria was read.

A reference management software was used for the control of the articles analyzed and for the removal of duplicates (Mendeley Desktop; Mendeley, London, UK). Disagreements were resolved by discussion between the authors. The histological/immunohistochemical description of the lesions reported in the studies was thoroughly assessed by three of the authors in order to confirm the diagnosis of odontoma (G.M.F., J.C.P. and D.R.M.F.A.).

The inclusion criteria were: (1) cases of odontoma or AFO of the jaws with sufficient histological and immunohistochemical data; (2) cross-sectional studies, in vitro studies, and case reports. The WHO classifies AFOs as the early developmental stages of odontomas [5]. Excluded were: (1) cases of odontoma in animals; (2) studies without the minimal immunohistochemical panel to support the diagnosis; (3) cases of odontoma associated with other odontogenic lesions; (4) cases of AFOs associated with other odontogenic lesions; (5) cases of ameloblastic fibroma and ameloblastic fibrodentinoma, and (6) reviews.

Data Extraction and Analysis

The reviewers independently screened the articles for data extraction. Any disagreements were resolved by discussion and Cohen’s kappa agreement among examiners was 0.80. The following data were extracted, when available: (1) author; (2) year; (3) location; (4) study design; (5) histopathological features; (6) immunohistochemical stains performed; (7) outcome.

Results

Study Selection

Using the search strategy developed in this systematic review, a total of 4426 studies were retrieved from the databases. After reading the titles and abstracts, 49 articles were considered potentially eligible and the full text was read by two evaluators (G.M.F., J.C.P. and D.R.M.F.A.). After analysis, 30 articles met all inclusion criteria and were selected for this systematic review, totaling 355 cases of odontomas and 43 cases of AFO. The articles are listed in the Supplemental Material.

Twenty-nine articles were published in English and one in Japanese. The countries where the studies were conducted were Japan (n = 19), Brazil (n = 6), Canada (n = 2), France (n = 1), South Korea (n = 1), and Mexico (n = 1). The publication date ranged from 1991 to 2019. The study designs were cross-sectional and in vitro studies, as well as case reports involving immunohistochemical methods. Thee cross-sectional studies had a STROBE score ≥ 14 and were included in this systematic review [8], Table S1. The flow diagram in Fig. 1 illustrates the screening and selection process of the articles. Ethics approval or patient consent was not required for this study.

Fig. 1.

The flow diagram illustrates the screening and selection of the articles following PRISMA guidelines

Histopathological and Immunohistochemical Characteristics

Odontomas

The histopathological types analyzed were compound odontomas (n = 210; 59.1%), with the largest number of cases, and complex odontomas (n = 111; 31.2%). Information about the histopathological subtype was absent in five articles and incomplete in three (Table 1). Ghost cells were also found in 79 cases of compound odontomas and in 18 cases of complex odontomas [9–11].

Table 1.

Systematic literature review of the main proteins expressed in odontomas

Author, year	Country	Study design	n	Histopathology	Antibodies	Intensity	Site	Outcome
Fujii et al. [13]	Japan	Cross-sectional and in vitro	21	Compound (15) Complex (6)	ß-Catenin	Strong	Odontogenic epithelium	Wnt/β-catenin pathway suppresses the proliferation of odontogenic epithelial cells
					Lef-1	Strong
Trejo-Remigio et al. [12]	Mexico	In vitro	7	Compound (4) Complex (3)	Amelogenin	Strong	Ectomesenchyme	Higher amelogenin expression in compound odontomas is related to the enamel knot, late-stage odontogenesis and ectomesenchymal interactions. On the other hand, the expression of CD34, SOX2 and OPN in complex odontoma could be responsible for the different behavior and mineralized amorphous structure
					BSP	Strong
					Pax9	Strong
					EDAR	Strong
					Barx	Strong
					Msx2	Strong
					Sox2	Strong
					CD34	Strong
					RUNX2	Strong
					OPN	Strong
Chau et al. [14]	Canada	Cross-sectional	10	Compound (2) Complex (8)	Periostin	Strong	Epithelium and ectomesenchyme	Periostin expressed in stroma is involved in the formation of enamel
Kiyoshima et al. [15]	Japan	Cross-sectional	11	Compound (8) Complex (3)	Thymosin β4	Negative	Enamel matrix	Thymosin β4 intensity was much higher in enamel matrix than in ameloblast-like cells. These findings were observed in both subtypes of odontoma
					Sheathlin	Strong
					Amelogenin	Strong
					Enamelin	Strong
Crivelini et al. [11]	Brazil	Cross-sectional	2	Complex (1)	Amelogenin	Strong	Odontogenic epithelium	Odontomas exhibited specialized secretory activity of ameloblasts that was not found in ameloblastic fibromas
					Sheathlin	Moderate
					Amelotin	Strong
					ODAM	Weak
Gonzales-Alva et al. [10]	Japan	Cross-sectional	86	Compound (57) Complex (29)	Podoplanin	Strong	Pulp	Podoplanin may be involved in the differentiation of pulp cells into odontoblasts in odontomas
Kim et al. [16]	South Korea	In vitro	4	Compound (3) Complex (1)	LHX8	Strong	Ectomesenchyme	LHX8 controls transcription factor families during tooth morphogenesis (BMPs, FGFs, SHHs, WNTs)
Fujita et al. [17]	Japan	Cross-sectional	39	Compound (24) Complex (15)	Midkine	Weak	Odontogenic epithelium	MK mediates cell growth and differentiation of odontogenic mixed tumors. Odontomas contain cells that are not fully differentiated
					Ki-67	Negative
Tanaka et al. [9]	Japan	Cross-sectional	69	Compound (52) Complex (17)	ß-Catenin	Strong	Odontogenic epithelium	The Wnt signaling pathway may be involved in the formation of ghost cells and accumulation of hair proteins (PA-HP1, PA-HP 2) in odontomas
					Lef-1	Strong
					Hair proteins 1,2	Strong
Fujita et al. [18]	Japan	Cross-sectional	62	Compound (40) Complex (22)	Vimentin	Strong	Pulp	Nestin is involved in the differentiation of pulp cells into odontoblasts in odontogenic tumors
					S100	Negative
					Nestin	Strong
Fujii et al. [19]	Japan	Cross-sectional	3	Not informed	Collagen IV	Strong	Ectomesenchyme and pulp	Differentiation and induction of tooth papilla
Crivelini et al. [20]	Brazil	Cross-sectional	3	Compound (3)	Vimentin	Negative	Odontogenic epithelium	CK 19 did not replace CK14 in secretory ameloblasts. Differentiation of ameloblasts seldomly occurs in odontomas
					CK7, CK14	Strong
					CK13, CK19	Negative
Abiko et al. [21]	Japan	Cross-sectional	2	Complex (1)	Amelogenin	Strong	Enamel matrix	Amelogenins are localized in secretory ameloblasts, a maturation stage of ameloblasts, and in the enamel matrix
So et al. [22]	Canada	Cross-sectional	3	Not reported	FGF1	Weak	Odontogenic epithelium	FGF-2 is involved in the histodifferentiation stage of odontogenesis
					FGF2	Moderate
Takata et al. [23]	Japan	Cross-sectional	10	Not reported	Amelogenin	Strong	Enamel matrix	Tumor cells differentiate into the stage of functional ameloblasts and secrete sheath proteins in the hyaline material
					Sheathlin	Strong
Takata et al. [24]	Japan	Cross-sectional	10	Not reported	Amelogenin	Strong	Enamel matrix	Enamelysin is a marker for the identification of secretory ameloblast-like cells
					Enamelysin	Strong
Papagerakisl et al. [25]	France	Cross-sectional	4	Complex (4)	Amelogenin	Strong	Epithelium and ectomesenchyme	Epithelial tumor cells are recapitulating genetic programs expressed during normal odontogenesis
					Osteocalcin	Strong
					Collagen III/IV	Strong
					MIB1	Negative
					Cytokeratin	Strong
					Vimentin	Strong
					Ki-67	Negative
Gao et al. [26]	Japan	Cross-sectional	2	Compound (2)	BMP	Strong	Odontogenic epithelium	Odontogenic tumors with formation of enamel, dentin, cementum or bone
Mori et al. [27]	Japan	Cross-sectional	5	Complex (1)	Amelogenin	Strong	Pulp	Pathological alteration in the epithelial ectomesenchymal interaction
					Tenascin	Strong
Mori et al. [28]	Japan	Cross-sectional	2	Not reported	Amelogenin	Strong	Enamel matrix	Reduced ameloblasts in the odontoma displayed the strongest staining for amelogenins

The antibodies most frequently used for the identification of enamel matrix proteins in odontomas was amelogenin in nine studies and sheathlin in three studies. Cytokeratins, vimentin and Ki-67 were analyzed in four studies. Beta-catenin, Lef1, tenascin, MIB1, collagen IV, S100, and podoplanin were used in two articles. Only one article reported the use of BSP, Pax9, Barx, Msx2, Sox2, CD34, RUNX2, OPN, osteocalcin, FGF1 and 2, enamelysin (MMP20), enamelin, PRKAR1A, periostin, amelotin, ODAM, LHX8, nestin, midkine, and BMP.

The immunostaining intensity of the antibodies used in odontomas is described in Table 1. Strong immunostaining was the most frequent (n = 30). Moderate intensity was reported for two antibodies (sheathlin and FGF2), weak intensity for three antibodies (ODAM, midkine, and FGF1), and seven antibodies were negative in odontomas (Thymosin β4, Ki-67, S100, CK13, CK19, vimentin, and MIB1). Regarding the site of immunostaining analyzed, the odontogenic epithelium was the most frequent in the studies (n = 9), followed by analysis of the enamel matrix (n = 5), ectomesenchyme (n = 5), and pulp (n = 4).

The positive and negative immunoexpression of the antibodies according to odontoma subtypes was reported in five articles. All studies using amelogenin showed strong immunostaining, especially in the ectomesenchyme and enamel matrix of compound odontomas [12]. Another important finding was the higher immunoexpression of Sox 2 in complex odontomas [12], Table 2.

Table 2.

Frequency of strong immunolabeling for different antibodies between complex and compound odontomas

Study	Antibody	Complex odontomas	Compound odontomas
Fujii et al. [13]	ß-catenin (nuclear)	3/6	12/15
Gonzáles-Alva et al. [10]	Podoplanin	16/29	41/57
Fujita et al. [17]	Midkine	2/15	1/24
Tanaka et al. [9]	Hair proteins 1,2	5/17	41/52
Fujita et al. [18]	Nestin	16/22	33/40

Ameloblastic Fibro-Odontomas

Cases of AFO corresponded to 10.8% of the sample (n = 43). Amelogenin was the most frequent antibody analyzed in six articles and was strongly expressed in the enamel matrix but absent in the epithelium. Sheathlin was another that was positive in the enamel matrix and negative in the epithelium.

The immunostaining intensity of the antibodies used in AFO is described in Table 3. Strong immunostaining intensity was the most frequent (n = 25). Moderate intensity was observed for three of the antibodies analyzed (DNMTs1, Thymosin ß4, and PRKAR1A) and weak intensity for 10 antibodies (Ki-67, S100, BrdU, CK8, CK14, CK18, periostin, FGF1, MIB1, and vimentin). Twelve antibodies were negative in the lesions studied (GFAP, CK4, CK7, CK13, sheathlin, amelogenin, enamelin, DNMTs3A, MIB1, Ki-67, tenascin, and S100). The odontogenic epithelium was the most frequently analyzed site in the studies (n = 14), followed by the ectomesenchyme (n = 9) and enamel matrix (n = 2).

Table 3.

Systematic literature review of the main proteins expressed in ameloblastic fibro-odontomas

Author, year	Country	Study design	n	Antibodies	Intensity	Site	Outcome
Lopes et al. [29]	Brazil	Case report	1	AE1/AE3	Strong	Epithelium and ectomesenchyme	CK14 and AE1/AE3 staining was strong, suggesting an early stage of the differentiation process
				CK14	Strong
				CK19	Strong
				Vimentin	Strong
				ß-catenin	Strong
				Ki-67	Weak
				S100	Weak
Sukegawa et al. [30]	Japan	Case report	1	WNT1	Strong	Epithelium and ectomesenchyme	ß-Catenin nuclear translocation was observed in stellate reticulum cells and in many papilla-like mesenchymal cells. Wnt1 and β-catenin were mainly present in the tumor nests
				ß-Catenin	Strong
Souza et al. [31]	Brazil	Cross-sectional	4	PRKAR1A	Moderate	Epithelium and ectomesenchyme	The PRKAR1A gene is expressed during normal tooth development and is mutated in ectomesenchymal tumors
Guimarães et al. [32]	Brazil	Cross-sectional	2	DNMTs1	Moderate	Epithelium and ectomesenchyme	Expression of DNA methyltransferases (DNMTs) 1 and 3B suggests methylation maintenance
				DNMTs 3A	Negative
				DNMTs 3B	Strong
Caetano et al. [33]	Brazil	Cross-sectional	4	Podoplanin	Strong	Odontogenic epithelium	Secretory ameloblasts expressed podoplanin, while mature ameloblasts did not
Chau et al. [14]	Canada	Cross-sectional	8	Periostin	Weak	Ectomesenchyme	The intermediately differentiated tumor would show intermediate staining
Kiyoshima et al. [15]	Japan	Cross-sectional	2	Thymosin β4	Moderate	Epithelium	Thymosin β4 might be associated with morphogenesis and tumor invasion
				Sheathlin	Negative
				Amelogenin	Negative
				Enamelin	Negative
So et al. [22]	Canada	Cross-sectional	3	FGF1	Weak	Odontogenic epithelium	FGF-2 is involved in nuclear activation during the histodifferentiation stage
				FGF2	Strong
Yagishita et al. [34]	Japan	Case report	1	Amelogenin	Strong	Epithelium	Amelogenin was detected almost exclusively in the induced enamel and dentinoid areas
Takata et al. [23]	Japan	Cross-sectional	4	Amelogenin	Strong	Enamel matrix	The areas of inductive hard tissue formation were stained with sheathlin
				Sheathlin	Strong
Takata et al. [24]	Japan	Cross-sectional	4	Amelogenin	Strong	Enamel matrix	Areas of inductive hard tissue formation and immature enamel were intensely stained, while the dentinoid material was not
				Enamelysin	Strong
Papagerakisl et al. [25]	France	Cross-sectional	1	Amelogenin	Strong	Epithelium and ectomesenchyme	These proteins are responsible for extracellular matrix deposition and were increased in ameloblastic fibro-odontomas
				Osteocalcin	Strong
				Collagen III and IV	Strong
				MIB1	Negative
				Cytokeratin	Strong
				Vimentin	Strong
				Ki-67	Negative
Sano et al. [35]	Japan	Cross-sectional	2	MIB1	Weak	Epithelium and ectomesenchyme	MIB1 indicates less proliferative potential in ameloblastic fibro-odontomas
Sekine et al. [36]	Japan	Case report	1	BrdU	Weak	Epithelium and ectomesenchyme	The mesenchymal component was more proliferative than the epithelial component
				PCNA	Strong
Miyauchi et al. [37]	Japan	Case report	1	CK4,7,13	Negative	Epithelium	The neoplastic epithelial cells of the present case seem to show cell differentiation corresponding to the bell stage
				CK8	Weak
				CK14,18	Weak
				CK16, 19	Strong
				Vimentin	Weak
				GFAP	Negative
Mori et al. [27]	Japan	Cross-sectional	2	Amelogenin	Strong	Epithelium and ectomesenchyme	The basement membrane of odontogenic epithelium and mesenchyme was diffusely positive
				Tenascin	Strong
Yamamoto et al. [38]	Japan	Cross-sectional	2	Tenascin	Negative	Odontogenic epithelium	Only immature dental papilla-like ectomesenchymal tissue was positive for tenascin
				Cytokeratin	Strong
				Vimentin	Weak
				S100	Negative
				Ki-67	Negative

Odontogenesis

The positive control group consisted of tooth germs (human or rats) and postnatal human teeth and was present in eight studies (Table 4). Differences in immunostaining were observed between the different stages of odontogenesis. The early bud and cap stages were characterized by strong immunoexpression of three antibodies (ß-catenin, Lef1, and BMP). Vimentin staining was weak in the early bell stage and negative in the other stages, while sheathlin and enamelysin exhibited weak immunostaining in the late bell stage. Strong immunostaining of amelogenins was observed during enamel formation and cytokeratins were strongly expressed during amelogenesis. CK13 and 19 are usually positive in odontogenesis. The main antibodies that differed between odontogenesis and tumor lesions were those against enamel matrix proteins: amelogenin, sheathlin, amelotin, ODAM, and enamelysin. These proteins exhibited low or no immunoexpression in odontoblasts, pre-dentin, pulp, and dental follicle.

Table 4.

Systematic literature review of the main proteins expressed during odontogenesis

Author, year	Type	Study design	n	Antibodies	Intensity	Stage of odontogenesis	Outcome
Fujii et al. [13]	Mice	Cross-sectional and in vitro	3	ß-Catenin	Strong	Bud and late cap stage	Activation of Wnt signaling disrupted tooth germ development by decreasing Sema3A
				Lef-1	Strong
Kim et al. [16]	Human	In vitro	5	LHX8	Moderate	Postnatal	LHX8 (homeobox gene 8) is expressed only during the early stage of tooth development, and the expression is diminished in the late bell stage
Crivelini et al. [20]	Human	Cross-sectional	3	Vimentin	Negative	Early bell stage Amelogenesis	CK13 and 19 appear in squamous differentiation or epithelial cells near the surface epithelium
				CK7, CK14	Strong
				CK13, CK19	Strong
Abiko et al. [21]	Rat	Cross-sectional	1	Amelogenin	Moderate	Enamel formation	Not expressed in all stages of odontogenesis, only in the stages of enamel formation. Amelogenin indicates differentiation
Takata et al. [23]	Human	Cross-sectional	5	Amelogenin	Moderate	Enamel formation Late bell stage	Sheathlin is a marker of functional differentiation of secretory ameloblasts and enamel matrix
				Sheathlin	Weak
Takata et al. [24]	Human	Cross-sectional	5	Amelogenin	Strong	Enamel formation Late bell stage	In tooth germs, enamelysin expression was detected only in the secretory enamel
				Enamelysin	Weak
Papagerakisl et al. [25]	Human	Cross-sectional	18	Amelogenin	Strong	Postnatal	Mineralized teeth express specific genes, such as those encoding collagens, osteocalcin, amelogenins, ameloblastin, and enamelins
				Osteocalcin	Strong
				Collagen III and IV	Strong
				Cytokeratin	Strong
				Vimentin	Strong
Gao et al. [26]	Human	Cross-sectional	2	BMP	Strong	Bud and cap stage	Plays an important role in the formation of enamel, dentin, cementum or bone

Outcome

Odontomas exhibited immunoexpression similar to that of control teeth, except for the immunoexpression of enamel matrix proteins which was lower in the positive controls. In summary, higher immunoexpression scores of proteins involved in cellular histodifferentiation were observed in odontomas, especially compound odontomas, Fig. 2.

Fig. 2.

Morphological features of different odontogenic mixed tumors and odontogenesis. a Tooth germ of mice containing enamel, dentin, and pulp tissue with odontoblasts (magnification: ×100, H/E). b AFOs showing mixtures of enamel, dentin and pulp tissues in a relatively loose stroma containing remnants of ameloblastomatous epithelium (magnification: ×50, H/E). c Complex odontomas showing mixtures of disorganized dental tissues (magnification: ×100, H/E). d Compound odontomas showing organized dental tissues that resembles a tooth (magnification: ×400, H/E)

Discussion

During odontogenesis, reciprocal signaling takes place between epithelial and ectomesenchymal tissues and the Wingless (Wnt)/β-catenin signaling pathway is essential for the early activation of odontogenesis. In contrast, aberrant activation of this signaling pathway has been associated with the formation of odontomas and ghost cells [9, 13].

The Wnt/β-catenin pathway and its target gene lymphoid enhancer-binding factor 1 (Lef1) have been reported to be involved in the expression of high-molecular weight cytokeratins [39]. The immunohistochemical findings of odontomas were similar to those observed during odontogenesis, especially those related to CK7 and 14. CK14 is a typical intermediate filament of odontogenic epithelium and its replacement with CK19 suggests advanced amelogenesis as a consequence of cellular secretory activity. However, it does not occur in odontomas, CK19 does not replace CK14 in secretory ameloblasts in advanced stages of amelogenesis; thus, differentiation of ameloblasts is not completed in odontomas, a fact that explains why CK19 is negative in odontomas and why the enamel matrix that does not undergo complete mineralization [20]. The exclusive immunoexpression of CK13 in tooth germs confirms squamous epithelial differentiation and the presence of epithelial cells near the surface [20].

Interestingly, CK14 was weakly immunoexpressed in the epithelium and strongly immunoexpressed in the ectomesenchyme of AFOs. In addition, CK19 showed strong immunostaining scores in the epithelium of AFOs similarly to that found in odontogenesis [20, 29, 37]. This finding agrees with the view that AFO would be a separate entity in odontogenesis and odontoma or can CK19 revert back to CK14. Further studies evaluating these proteins are necessary for a better understanding.

Vimentin is a member of the intermediate filament family which, together with microtubules and actin microfilaments, makes up the cytoskeleton that is responsible for the maintenance and integrity of the cytoplasm of cells. This protein has been used as an immunohistochemical marker to identify mesenchyme-derived tissues [40]. The present review showed that immunohistochemical staining for vimentin occurs in stages prior to the bell stage during odontogenesis and is negative in subsequent stages. Positive staining was observed in AFOs, while odontomas were negative for this protein. These findings demonstrate that the expression pattern of molecular markers in odontomas is very similar to that observed during odontogenesis and reinforce that AFOs contain earlier tissues of odontogenesis [3, 41].

The genes encoding enamel proteins also possess binding sites in the Lef1 promoter domain [42] and are reported to be essential for formation of the enamel matrix. These proteins are divided into two classes, amelogenins that are soluble in salts and non-amelogenins, which include enamelins, enamelysins, sheathlin (ameloblastin), amelotins, and odontogenic ameloblast-associated protein (ODAM) that is found close to hydroxyapatite crystals [11]. In young or immature enamel, antibodies against amelogenins stain 90–95% of enamel proteins, while antibodies against enamelins stain only 5–10%. Amelogenins and enamelins are biosynthesized by young ameloblasts and are secreted within the extracellular matrix of enamel [28]. Strong staining intensity for amelogenin was reported in seven studies [11, 21, 23–25, 27, 28]. During tooth development, amelogenin is expressed mainly in the bell stage, indicating differentiation [21], while the high immunoexpression of amelogenins in odontomas is due to the high specialized secretory activity of ameloblasts [11].

The antibody used for the identification of ameloblastin (sheathlin) is considered a good marker of the functional secretory differentiation of ameloblasts and of the enamel matrix. This protein was found to be weakly expressed during the bell stage of odontogenesis [23]. On the other hand, moderate [11] and intense expression was observed in odontomas [15, 23], as tumor cells differentiate into functional ameloblasts and secrete proteins into the matrix [23]. Enamelysin (MMP20) is another immunomarker of secretory ameloblast-like cells that exhibited weak reactivity in tooth germs and intense reactivity in odontomas [24].

Thymosin β4 plays an important role in cell motility by regulating actin polymerization and might be associated with morphogenesis and tumor invasion. This protein showed moderate and negative immunoexpression in AFOs and odontomas, respectively [15]. This result thus confirms the indolent behavior of these tumors.

PRKAR1A is a tumor suppressor gene that encodes protein kinase A. Alterations in this gene lead to an increase in PKA/cAMP signaling and activation of the proliferation and differentiation process in tumors of ectomesenchymal origin. This gene is mutated in AFOs [31]. Further studies are needed to evaluate whether odontomas carry the same PRKAR1A gene mutation as AFOs. This may be a possible area of future research to confirm that AFOs are immature odontomas.

MIB1, together with the Ki-67 proliferative index, is used to evaluate the growth potential of tumors. Staining for this protein was weak and/or negative in odontomas and AFOs [25, 35]. This finding indicates that these tumors have a higher degree of development and are less proliferative than other mixed odontogenic tumors such as ameloblastic fibromas [35].

Fibroblastic growth factor (FGF) plays important roles in various stages of odontogenesis and in the signaling process between the epithelium and ectomesenchyme. FGF-2 is the subtype involved in the stages of histodifferentiation during odontogenesis. This protein was moderately expressed in the epithelium of odontomas and strongly expressed in the epithelium of AFOs [22]. This difference is due to the greater nuclear activation during the process of histodifferentiation, which is more intense in the exchange of signals between the epithelium and stroma in AFOs.

Tenascin is a multifunctional glycoprotein involved in cell–cell and cell-extracellular matrix interactions. It is expressed at the epithelial-ectomesenchymal interface during embryo development and can cause pathological changes in epithelial-ectomesenchymal interactions in odontomas [27]. Strong immunoexpression of tenascin was reported in the pulp of odontomas and in the ectomesenchyme of AFOs [27]; however, tenascin was negative in the odontogenic epithelium of AFOs [38]. The explanation for this finding is that only dental papilla-like ectomesenchymal tissues were positive for tenascin [38]. Taken together, the data suggest that the accumulation of tenascin during odontogenesis is crucial for the development of mixed odontogenic tumors and that the dental papilla is the environment that provides the nutrients and adhesion molecules necessary for cell–cell and cell-extracellular matrix interactions.

Podoplanin also exhibited strong immunostaining intensity in the odontogenic epithelium of AFOs and odontomas as it is expressed in secretory ameloblasts but not in mature ameloblasts [10, 33]. The differentiation of pulp cells into odontoblasts in odontomas was demonstrated by podoplanin [10] and nestin [18].

Periostin regulates cell-extracellular matrix interactions and is involved in the formation of enamel [14]. This protein was highly expressed in the stroma of odontomas compared to AFOs, while similar immunoexpression was observed in the odontogenic epithelium of the two lesions [14]. Additionally, FGF2 protein was more immunoexpressed in AFOs. These findings suggest similar immunoexpression of these proteins in odontomas and AFOs.

Comparison of the histological types of odontoma showed that most proteins involved in the differentiation of dental tissues were more immunoexpressed in compound odontomas, including podoplanin, nestin, beta-catenin, and hair proteins 1,2 (composed of hard keratins and matrix proteins associated with Wnt-ß-catenin/Lef1 pathway mutation) [9]. In contrast, midkine was more expressed in complex odontomas. The explanation for this difference is that compound odontomas require more proteins involved in cell differentiation for tooth development, while the detection of midkine in complex odontomas suggests the presence of not fully differentiated cells [17], such as stem cells immunostained by the Sox2 protein in complex odontomas [12].

Comparison of odontomas and tooth germs showed higher immunostaining scores for proteins involved in enamel formation (amelogenin, sheathlin, enamelysin) and histodifferentiation (ß-catenin nuclear) had a higher propensity to development odontomas.

Odontomas are one of the most frequent causes of eruption disturbances. The scarcity of published studies in the literature analyzing proteins in odontomas due to the difficulty in manipulating the techniques for descaling this lesion was the main limitation of this systematic review. Thus, further research is needed to understand the development of odontomas and the mechanisms that differentiate them from odontogenesis.

Conclusions

The present systematic literature review showed that compound odontomas exhibit the highest immunoexpression of proteins involved in histodifferentiation in the odontogenic epithelium and/or enamel matrix, with the Wnt/beta-catenin pathway being involved in tumor formation of tumor. In AFO, immunoexpression of mutated proteins involved in proliferation such as the PRKAR1A gene is observed, as well as immunoexpression of proteins involved in differentiation similar to that seen in odontomas. Enamel matrix proteins are differently expressed in odontomas and tooth germs, with the higher expression of these proteins being related to the development of odontogenic tumors such as odontomas.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Funding

None.

Compliance with Ethical Standards

Conflict of interest

The authors declare that they have no competing interests.