Tetracyclines and bone: unclear actions with potentially lasting effects

Abstract

Tetracyclines are a broad-spectrum class of antibiotics that have unclear actions with potentially lasting effects on bone metabolism. Initially isolated from Streptomyces, tetracycline proved to be an effective treatment for Gram +/− infections. The emergence of resistant bacterial strains commanded the development of later generation agents, including minocycline, doxycycline, tigecycline, sarecycline, omadacycline, and eravacycline. In 1957, it was realized that tetracyclines act as bone fluorochrome labels due to their high affinity for the bone mineral matrix. Over the course of the next decade, researchers discerned that these compounds are retained in the bone matrix at high levels after the termination of antibiotic therapy. Studies during this period provided evidence that tetracyclines could disrupt prenatal and early postnatal skeletal development. Currently, tetracyclines are most commonly prescribed as a long-term systemic therapy for the treatment of acne in healthy adolescents and young adults. Surprisingly, the impact of tetracyclines on physiologic bone modeling/remodeling is largely unknown. This article provides an overview of the pharmacology of tetracycline drugs, summarizes current knowledge about the impact of these agents on skeletal development and homeostasis, and reviews prior work targeting tetracyclines’ effects on bone cell physiology. The need for future research to elucidate unclear effects of tetracyclines on the skeleton is addressed, including drug retention/release mechanisms from the bone matrix, signaling mechanisms at bone cells, the impact of newer third generation tetracycline antibiotics, and the role of the gut-bone axis.

1. Introduction

Tetracyclines are a broad-spectrum class of antibiotics that have unclear actions with potentially lasting effects on bone metabolism (Figure 1). Tetracyclines are used for the treatment of bacterial infections, periodontitis, and dermatological conditions.^1–6 Tetracycline drug-induced bone fluorochrome labeling and minocycline-induced “black bone” discoloration demonstrate that tetracyclines are retained in the bone matrix.^7–11 Tetracyclines have detrimental effects on prenatal and early postnatal skeletal development and can attenuate ovariectomy-induced osteopenia.^12–20 In vitro investigations have shown that tetracyclines impact osteoclastogenesis and osteoblastogenesis.^21–29

Figure 1. Tetracyclines and bone.

[Two-Column Fitting Image] Tetracyclines are used for clinical dynamic bone histomorphometry as well as for the treatment of bacterial infections, periodontitis, and dermatological conditions. Tetracycline antibiotic bone fluorochrome labeling and minocycline-induced black bone discoloration demonstrate that tetracyclines are retained in the bone matrix. Tetracyclines have detrimental effects on early skeletal development and can attenuate experimentally induced osteopenia. In vitro investigations have delineated that tetracyclines can suppress osteoclastogenesis and suppress or promote osteoblastogenesis.

2. Pharmacology of Commonly Administered Tetracyclines

Tetracyclines are a class of antibiotics that were first discovered in 1948 when Duggar isolated chlortetracycline from the soil bacterium Streptomyces aureofaciens.³⁰ Soon thereafter, other natural tetracyclines produced by Streptomyces were also isolated, including oxytetracycline and tetracycline.² Tetracycline was favored clinically due to its higher potency, better solubility profile and more favorable pharmacological activity.¹

The tetracycline scaffold is based on a ABCD naphthacene ring system containing four aromatic rings (Figure 2A). Tetracycline antibiotics exert their antimicrobial activity through inhibition of bacterial protein synthesis by binding to the decoding center in the 30S ribosomal subunit, thus preventing the binding of aminoacyl-tRNA to the A site. Specifically, the conserved hydrophilic surface of tetracycline interacts with the irregular minor groove of helix 34 and loop of helix 31 of 16S rRNA.³¹ In this position, the C and D rings of tetracycline (Figure 2B) sterically hinder the interaction between the first nucleotide of the anticodon of the tRNA and the third base of the mRNA codon.³¹ Modifications to the ABCD naphthacene ring system alter the pharmacological profiles of tetracycline drugs.

Figure 2. Chemical structure of tetracycline drugs.

(A) ABCD napthacene core ring system (B) Naturally occurring tetracycline. (C) Doxycycline, an analog of tetracycline, with a rearrangement of the methyl group from position carbon 5 (C5) to position C6. (D) Minocycline, an analog of tetracycline, with a dimethylamino group at position C7 and lacking the methyl and hydroxy groups at position C5. (E) Tigecycline, an analog of minocycline, with a 9-tert-butyl-glycylamido side chain at position C9 on the D ring. (F) Omadacycline, an aminomethylcycline with structural modifications of minocycline at D9 with an aminomethyl group. (G) Eravacycline contains a tetracyclic core, with the addition of a fluorine at position C7 and a pyrrolidinoacetamo group at C9. (H) Sarecycline, another minocycline analog, but with a replacement of the dimethylamino group at position C.

The emergence of antimicrobial resistance to early tetracyclines spurred the development of later generation tetracycline drugs.^32,33 Doxycycline and minocycline are second generation, semisynthetic tetracycline antibiotics introduced in the 1960s.^1,34 Doxycycline (Figure 2C) was designed to increase drug stability, as well as oral availability with improved lipid solubility, by shifting the methyl group on tetracycline from ring B to ring C.^35,36 Compared to its tetracycline parent, minocycline (Figure 2D) contains an additional dimethylamido group on carbon 7 (C7) of the D ring and exhibits increased affinity to the rRNA target, through increased stacking interactions with helix 34.³¹ Tigecycline (Figure 2E), a third generation semisynthetic tetracycline drug, was introduced in 2005. Tigecycline exhibits a chemical modification to minocycline at C9 on the D ring.^37–39 Introduction of a 9-tert-butyl-glycylamido side chain enhances binding affinity by increasing stacking interactions with 16S rRNA.^31,40 In 2018, the FDA approved three new tetracycline drugs, including two semisynthetic agents, omadacycline and sarecycline, and the first fully synthetic compound, eravacycline.⁴¹ Omadacycline (Figure 2F) is a minocycline derivative with a C9 substitution on the D ring. The aminomethyl group substitution on ring D confers 2-fold higher affinity for the ribosome and inhibits in vitro translation at 2-fold lower drug concentrations compared to tetracycline.³¹ In addition, the C9 substitution increases oral bioavailability and lowers the adverse gastrointestinal effects associated with tigecycline.⁴⁰ Eravacycline (Figure 2G) contains a tetracycline core with modifications that include a fluorine atom at C7 and a pyrrolidinoacetamido group at C9 of the D ring. This confers a 10-fold higher affinity for the ribosome and inhibits in vitro translation at 4-fold lower drug concentrations compared to tetracycline.³¹ Sarecycline (Figure 2H) is a minocycline analog that has 7-[[methoxy(methyl)amino]methyl] group introduction at C7 on the D ring. The C7 moiety interacts directly with mRNA at the A-site codon. This interaction may be responsible for sarecycline’s reduced adverse side effects compared to other tetracycline drugs.⁴²

3. Clinical Applications of Tetracyclines

Tetracycline, doxycycline, minocycline, and tigecycline are the most commonly used tetracycline antibiotics when treating bacterial infections (Table 1)^43,44. Tigecycline use is limited to resistant infections that are refractory to earlier generation tetracyclines. Tigecycline use is restricted because an increase in all-cause mortality was observed among adult patients during clinical trials of the agent.^45,46 Omadacycline was introduced to treat adults with community-acquired bacterial pneumonia and acute bacterial skin and skin structure infections, including those resistant to earlier generation tetracyclines.^41,47 Eravacycline is used to treat complicated, antibiotic resistant intra-abdominal infections.^48,49 Sarecycline is FDA approved for the treatment of moderate to severe acne⁴¹ and was developed to limit adverse gastrointestinal and phototoxicity side effects caused by off-target actions of earlier generation tetracycline drugs.⁵⁰ The low incidence of adverse gastrointestinal effects with sarecycline is due to its reduced activity against Gram-negative bacteria and anaerobes present in the intestinal microbiota.⁵⁰

Table 1:

Clinical uses of tetracycline drugs

Antibiotic	Origin	Half-life elimination	Uses
Tetracycline	Natural51	6–11 hours52	Acne vulgaris53, sexually transmitted infections54, Vibrio cholerae55 Off-label·. Helicobacter pylori eradication56, periodontitis57
Minocycline	Semi-synthetic51	11–17 hours58	Acne vulgaris, Lyme disease, perioral infections Off-label: Gastrointestinal, sexually transmitted, soft tissue, Rickettsia, zoonotic infections59
Doxycycline	Semi-synthetic51	18–22 hours60	Acne vulgaris53, periodontitis57, sexually transmitted infections54 Off-label: Anthrax61, malaria62
Tigecycline	Semi-synthetic51	~40 hours63	Complicated skin and intra-abdominal infections64,65 Multidrug-resistant organisms37
Omadacycline	Semi-synthetic41	~16 hours47	Community-acquired bacterial pneumonia, Skin and skin structure infections47
Eravacycline	Synthetic41	20 hours48	Complicated intra-abdominal infections41
Sarecycline	Semi-synthetic66	21–22 hours67	Acne vulgaris41

In addition to bacterial infections, tetracyclines are also administered to treat dermatological conditions and periodontal disease.¹ Of these, the most common use of tetracycline drugs is for the treatment of acne vulgaris.^43,68,69 Acne afflicts approximately 50 million Americans annually, of whom roughly 85% are between the ages of 12 and 24.^70,71 Overall, about 1/3 of adolescents and young adults are prescribed oral antibiotics for the treatment of acne.^72,73 According to the 2016 Status Report from the Scientific Panel on Antibiotic Use in Dermatology of the American Acne and Rosacea Society, tetracyclines constitute 75% of oral antibiotics prescribed by dermatologists.⁷⁴ The pathophysiology of acne is mostly attributed to the opportunistic bacteria Cutinebacterium acnes (C. acnes), formerly known as Propionibacterium acnes.^70,71 Minocycline and doxycycline are administered to target C. acnes^75–77. Clinical trials have found that sarecycline has similar efficacy against C. acnes, but with reduced adverse gastrointestinal and phototoxicity side effects.^78–81 Although practice guidelines recommend a minimum of 6–8 weeks and a maximum of 3–6 months for administration of oral antibiotics for acne^82–84, more than 60% of oral antibiotic courses extend longer than three months, with eleven months being the average duration of treatment.^72,73 The standardized doses of oral tetracycline, doxycycline, minocycline, and sarecycline for the treatment of acne in children, adolescents, and adults are summarized in Table 2.

Table 2:

Standard oral dosing of tetracyclines for acne

Antibiotic	*Children ≥ 8 yro and adolescents	Adults
Tetracycline	500 mg twice daily44	1 g daily in divided doses; reduce gradually to 125 to 500 mg/day53
Doxycycline	50 to 100 mg once or twice daily or 150 mg once daily44	50 to 100 mg twice daily or 100 mg once daily53,85
Minocycline	50 to 100 mg once or twice daily44	50 to 100 mg twice daily53
Sarecycline	60 mg daily (33–54 kg patients) 100 mg daily (55–84 kg patients) 150 mg daily (85–136 kg patients)41  *Children ≥ 9 yro and adolescents	60 mg daily (33–54 kg patients) 100 mg daily (55–84 kg patients) 150 mg daily (85–136 kg patients)41

Administering tetracyclines at antimicrobial doses (Table 2) can result in a range of undesirable side-effects, including adverse gastrointestinal and hepatic reactions.^67,86 Dose-related gastrointestinal side effects, including abdominal discomfort, epigastric pain, nausea, vomiting, and diarrhea, are commonly reported by patients taking oral tetracyclines and intravenous tigecycline.⁶⁷ Minocycline has most frequently been linked drug-induced liver injury. Short-term use of minocycline may result in acute-hepatitis-like syndrome, while long-term use may cause chronic hepatitis with autoimmune features. These side-effects are associated with a hepatocellular pattern of elevated serum enzymes, autoantibodies, and immunological features such as fever, rash, and eosinophilia.⁸⁶ Although doxycycline and minocycline share a similar chemical structure, doxycycline is not linked to autoimmune-like hepatitis. Doxycycline-induced liver injury is commonly characterized as a mix between hepatocellular and cholestatic.⁸⁶ Tetracycline-induced liver injury is rare. Acute fatty liver disease has been reported in pregnant women when given tetracycline in high doses.⁸⁶ The hepatotoxicity caused by minocycline, doxycycline, and tetracycline typically resolves after terminating antibiotic use.^67,86

To reduce these adverse effects, tetracyclines have been administered at sub-antimicrobial dosing for acne treatment.⁸⁷ In 1998, the FDA approved Periostat (sub-antimicrobial 20 mg doxycycline hyclate capsule administered twice daily) as an adjunctive host immunomodulatory treatment for chronic periodontitis.^6,88–90 In 2006, the FDA approved Oracea as a sub-antimicrobial dose doxycycline treatment for rosacea, which is administered once daily.⁹¹ Oracea is a 40 mg modified-release doxycycline capsule, which consists of 30 mg immediate-release and a 10 mg delayed-release mechanism.^3,92–94

4. Retention of Tetracyclines in Bone

One of the more remarkable properties of tetracyclines is their retention in the bone matrix. This was first recorded in 1956 when it was observed that a single intravenous injection of tetracycline led to incorporation and retention of the antibiotic in the bone of young mice.⁷ Later, Milch et al. (1957) administered a single intraperitoneal injection of tetracycline, chlortetracycline, or oxytetracycline (0.1 to 200 mg/kg) to mice, rats, guinea pigs, rabbits, and dogs and observed a brilliant yellow-gold fluorescence under ultraviolet light in long bones and flat bones across all species.⁸ Fluorescence could be detected at low doses (0.3 mg/kg) and throughout the 10-week observation period.⁸ Milch et al. (1958) employed subcutaneous, intramuscular, and intravenous injections or oral administration to show that the mode of administration had no effect on tetracycline-induced bone fluorescence in both laboratory animals and human subjects.⁹ Administering a clinical dose of tetracycline (1–2 g/day) was sufficient to induce bone fluorescence in humans. Whereas sex had no effect, tetracycline-induced bone fluorescence was influenced by age. Labeling was more intense in bone specimens from young versus mature adult subjects and appeared to occur only at regions of active bone formation. Fluorescence was intensely localized to the trabecular bone secondary spongiosa and the endosteal surfaces, periosteal surfaces and haversian canals of cortical bone. Since tetracyclines chelate metallic ions, including calcium, the authors speculated that fluorescence resulted from binding of the tetracycline naphthacenecarboxamide nucleus to calcium in the bone. The authors highlighted the need for future research to elucidate the mechanism by which tetracyclines are incorporated in bone, to measure the half-life when bound to bone, and understand the biological implications for skeletal health.⁹

To address these questions, subsequent studies evaluated the metabolism of tetracycline, and determined the concentration and persistence of the drug retained in bone.^10,95 Kelly and Buyske (1960) and Buyske et al. (1960) showed about 90% of tetracycline was eliminated via the fecal or urinary route, while 3–6 % of the drug was retained by the skeleton in Sherman rats.^10,95 Tetracycline was retained in the femur after a single 250 mg/kg dose by intraperitoneal injection and did not begin to deplete until 5–7 months post administration. Prolonged treatment increases the retention of tetracycline in the skeleton, suggesting that bone could act as a drug “depot” for tetracyclines with potentially lasting effects.¹⁰ Myers and Jaffe (1965) also demonstrated that tetracycline is retained in the skeleton of Long-Evans rats, with up to 5% of tetracycline retained within the skeleton. An additional finding was that more tetracycline was retained in the bones of young versus mature rats.¹¹

Retention in bone led Frost and collaborators at the Henry Ford Hospital to pioneer the clinical use of tetracycline as a fluorochrome marker for bone.⁹⁶ Their work showed that independent of the route of administration, tetracycline was deposited in the hydroxyapatite of actively mineralizing bone, hyaline cartilage, and developing teeth.⁹⁶ Frost and colleagues utilized pulsed tetracycline labeling of bone to develop histomorphometric approaches to quantify physiologic rates of bone formation, mineralization, and remodeling.^97–100 These approaches were seminal in establishing standardized dynamic bone histomorphometry techniques,^101,102 that continue to be used in research and clinical care today. Tetracycline and doxycycline remain the most commonly used bone fluorochrome labels in humans.¹⁰³

Different from other tetracyclines, minocycline does not have fluorescent properties in bone.¹⁰³ However, minocycline has been associated with osseous tissue discoloration, providing indirect evidence of retention in the bone matrix.^104,105 This phenomenon of minocycline-induced bone discoloration is commonly referred to as “black bone disease”.^106–109 Case reports in the dental literature commonly document that patients with a history of long-term minocycline therapy present with black/blue pigmentation of the mucogingival tissues.^{105,109–115}. Surgical exploration has revealed that alveolar mucogingival tissues were not pigmented, but rather it was the underlying alveolar bone.^105,110 Minocycline-induced black/blue bone discoloration has also been observed at non-oral skeletal sites, including the clavicle, vertebrae, acromion, femur, tibia, metatarsals, and metacarpals. In all cases, patients had undergone long-term minocycline therapy for the treatment of acne or rosacea, spanning from 2 to 10 years.^{104,106–108,116–121} It is theorized that black/blue discoloration of bone is caused by the oxidation of minocycline and the subsequent incorporation into ossifying tissue.^{104,110,112,122} Patients are advised to discontinue minocycline treatment if they observe dark pigmentation of the oral mucosa, gingiva, and skin. However, terminating minocycline administration does not resolve the soft tissue pigmentation when the subjacent bone is discolored.^112,121 The impact of long-term minocycline retention on bone quality and remodeling is largely unknown.

5. Effects of Tetracyclines on Perinatal and Early Postnatal Skeletal Developmental

Since tetracyclines have affinity for the bone matrix, questions arose whether this would affect early skeletal development (Table 3). Bevelander et al. (1959, 1960) carried out a series of seminal studies evaluating the teratogenic effects of tetracycline by injecting the antibiotic into the yolk sac of eight-day old chick embryos. Whereas lower doses (0.1 to 0.5 mg) had only minimal effects, higher doses (2.5 to 5.0 mg) induced pronounced retardation in the overall growth of the embryos. Moreover, tetracycline stunted skeletal development and led to malformed bones, which was attributed to lower trabeculae numbers and reduced mineralization.^12,13 Hughes et al. (1965) later found that injection of chick embryos with a clinically relevant dose of tetracycline (2.5 μg/g) did not interfere with bone growth. Only at ten times higher doses (25 μg/g) was bone growth inhibition and bone malformation observed.¹²³

Table 3:

Tetracycline treatment effects on early skeletal development

Study	Model	Subject Age	Antibiotic Dose	Administration Mode	Treatment Duration	Outcomes
Bevelander et al. (1959,60)12,13	Chick embryo	8 days	2.5 or 5.0 mg	Yolk-sac injection	Single dose	Stunted, malformed bones ↓ Femoral trabeculae & mineralization
Hughes et al. (1965)123	Chick embryo	4 days	25 μg/g	Allantois injection	Single dose	↓ Bone growth
Bevelander & Cohlan (1962)14	Pregnant rats to assess fetal outcomes	-	40–80 mg/kg/day	Intramuscular injection	12–20, 8–15, or 10–15 gestation days	↓ Skeletal growth
Cohlan et al. (1963)15	Premature infants	1 to 36 days	14–50 mg/kg/day	Oral	9–30 days (4 groups with alternating durations and recovery periods)	↓ Bone growth
Porter et al. (1965)17	Pregnant women to assess postnatal outcomes	-	1 g/day	Oral	2 or 6 weeks, during 1st, 2nd or 3rd trimester	Yellow/brown discoloration and enamel hypoplasia in deciduous dentition
Genot et al. (1970)18	Pregnant women to assess postnatal outcomes	-	1 g/day	Oral	2 or 6 weeks, during 1st, 2nd or 3rd trimester	Yellow/brown discoloration and enamel hypoplasia in deciduous dentition
Wallman & Hilton (1962)131	Healthy Infants	1 day	21–29 mg/kg/day	Oral	3 to 6 days	Yellow/brown discoloration in deciduous dentition

There is also evidence that tetracyclines affect the skeletal development of mammals in utero. This is of particular concern because antibiotics, including tetracyclines, can traverse the placental barrier from mother to fetus during pregnancy.^124–126 To ascertain whether transplacental tetracycline impacts the early developing skeleton, Bevelander & Cohlan (1962) administered tetracycline to pregnant rats and found that the skeleton of embryos and newborn pups retained tetracycline and exhibited growth retardation.¹⁴ Tetracycline-induced skeletal fluorescence was observed in a nonviable premature infant, whose mother had been administered 1 gram of oral tetracycline daily for 3 weeks prior to delivery.¹⁵ This case demonstrated that tetracycline can pass through the placenta and be retained in the human fetal skeleton. To investigate how tetracycline affects fetal skeletal development, Cohlan et al. (1963) administered the antibiotic orally to healthy premature infants.¹⁵ Using an x-ray technique to measure fibula growth rate, it was shown that tetracycline caused a dose-dependent inhibition of longitudinal bone growth (14–50 mg/kg/day) by up to 40%. Inhibitory effects on bone growth were reversible in groups in which antibiotic therapy was discontinued after a short period of administration.¹⁵ Two double-blind, placebo-controlled clinical studies published in 1965¹⁷ and 1970¹⁸ reported the impact of prenatal tetracycline exposure on skeletal growth and development. Bone development was similar in children whose mothers had taken tetracycline during pregnancy in comparison to those who had not. However, tetracycline exposure during the second and third trimesters caused high incidence of yellow/brown discoloration and enamel hypoplasia in the deciduous teeth.^17,18 These findings support the view that short-term use of tetracycline during pregnancy poses minimal risk for fetal skeletal development, but that long-term use is contraindicated.^127,128 As a result, tetracyclines have been classified as category D drugs. This means that there is evidence of risk to the human fetus, but that benefits from use in pregnant women may outweigh the risk. For this reason, tetracyclines are not recommended as a first-line antimicrobial treatment for pregnant women.^67,129

There is also evidence that tetracycline can be excreted in breast milk from mother to offspring. For instance, administering tetracycline to lactating female rats for three weeks following birth resulted in tetracycline labeling in the teeth and skeletons of their offspring.¹²⁴ As tetracyclines are lipid-soluble drugs, they can diffuse readily through lipid-filled membranes such as those in the mammary glands. Although there have been no reports of adverse effects to nursing infants, clinical studies have shown that tetracyclines are present at low concentrations in the breast milk of mothers receiving tetracycline therapy post-partum.¹³⁰

The risk of tetracyclines given directly to children must also be considered. Wallman and Hilton (1962) examined 50 children who had been given a course of tetracycline following birth.¹³¹ Typically, the tetracycline treatment occurred during the first week of postnatal life. Forty-six of 50 children had yellow/brown discoloration of the deciduous dentition, a characteristic of tetracycline-induced tooth staining.¹³¹ Furthermore, studies have shown that administering tetracycline to pediatric patients younger than eight years old results in high incidence of tetracycline-induced tooth staining. This occurs because the mineralization of the succedaneous dentition does not reach completion until a child reaches eight years of age (excluding third molars).^132,133 For this reason, clinical guidelines advise that tetracyclines should not be administered to children younger than eight years old.⁶⁷

6. Impact of Tetracyclines on Osteopenia

6.1. Post-menopausal osteopenia

Tetracycline drugs affinity for the bone matrix further led investigators to study these agents as a therapeutic intervention for post-menopausal bone loss (Table 4). Utilizing the ovariectomized (OVX) rat model, Williams et al. (1996) showed that oral minocycline administration (10 mg/day) improved bone mineral density and trabecular bone microarchitecture properties of the femur. The protective effect of minocycline against osteopenia was attributed to an increase in bone formation, with concurrent decrease in bone resorption.¹⁶ Li et al. (2003) found that oral administration of tetracycline (1.2 or 4.8 mg/kg/day) decreased bone resorption and increased osteogenesis in OVX rats. Interestingly, the 1.2 mg/kg/day dose increased osteoblast recruitment, while the higher dose enhanced osteoblast activity.¹⁹ Pytlik et al. (2004) later found that doxycycline (20 mg/kg/day) improved OVX-induced impaired bone mechanical properties,²⁰ and de Figueiredo et al. (2019) showed that doxycycline treatment (10 or 30 mg/kg/day) ameliorated OVX-induced osteopenia by improving cancellous bone microarchitecture, increasing bone mineral density, and reducing osteoclastogenesis.¹³⁴ Together, these findings indicate that tetracyclines have a protective effect on the skeleton in the post-menopausal state.

Table 4:

Tetracycline, minocycline, and doxycycline treatment effects on osteopenia

Study	Model	Sex	Experimental Osteopenia	Antibiotic (Dose)	Administration Mode	Treatment Duration	Outcomes
Williams et al. (1996)16	Wister rats	F	OVX at age 22–24 months	Minocycline (10 mg/d)	Oral gavage	8 weeks, initiated at OVX	↑ Femur bone mineral density ↑ Femur trabecular bone area, number, and thickness ↑ Bone formation ↓ Bone resorption
Li et al. (2003)19	Sprague Dawley rats	F	OVX at age 3.5 months	Tetracycline (1.2 or 4.8 mg/kg/d)	Chow	90 days, initiated at OVX	↑ Trabecular bone area ↑ Osteogenesis ↓ Bone resorption
Pytlik et al. (2004)20	Wister rats	F	OVX at age 3 months	Doxycycline (20 mg/kg/d)	Oral gavage	4 weeks, initiated 3–4 days after OVX	↑ Bone mineral content / bone mass ratio in femur Improved mechanical properties of the femur
de Figueiredo et al. (2019)134	Wister rats	F	OVX at age 2.5 months	Doxycycline (10 or 30 mg/kg/d)	Drinking water	60 days, initiated 3 months after OVX	↑ Bone mineral density in femur induced by 10 mg/kg/d ↑ Trabecular bone in femur induced by 10 & 30 mg/kg/d ↓ Osteoclasts induced by 30 mg/kg/d
Golub et al. (1990)135	Sprague Dawley rats	M	STZ-induced diabetes in adult rats	Doxycycline (2 mg/d)	Oral gavage	6 or 14 weeks	↑ Femur mineral content (calcium, phosphorus) ↓ Urinary calcium excretion
Fowlkes et al. (2015)136	DBA/2J mice	M	STZ-induced diabetes at age 10–11 weeks	Doxycycline (28–92 mg/kg/d)	Chow	10 weeks, initiated at age 12–13 weeks	No effect on femur trabecular / cortical bone microarchitecture No effect on femur biomechanical properties
Bain et al. (1997)137	Sprague Dawley rats	M	STZ-induced diabetes at age 3 months	Minocycline (20 mg/d)	Oral gavage	26 days	↑ Trabecular bone area in tibia ↑ Bone formation and mineral apposition rates in tibia
Sasaki et al. (1991,1992) 138,139	Sprague Dawley rats	M	STZ-induced diabetes at age 4 months	Minocycline (20 mg/d)	Oral gavage	21 days	↑ Osteoblast morphology and function
Kaneko et al. (1990)140	Sprague Dawley rats	M	STZ-induced diabetes in adult rats	Minocycline (20 mg/d)	Oral gavage	21 days	No effect on osteoclast structure and function

6.2. Diabetes-induced osteopenia

Similarly, tetracyclines’ chelating properties in bone led researchers to evaluate the drugs as a therapeutic intervention for diabetes-induced osteopenia (Table 4). In an observational study utilizing doxycycline, Golub et al. (1990) reported that oral doxycycline administration appeared to prevent osteopenia in streptozotocin (STZ)-induced diabetic rats (2 mg/day).¹³⁵ Contrasting these findings, Fowlkes et al. (2015) observed that orally administering doxycycline (28–92 mg/kg/day) to diabetic DBA/2J mice did not prevent or alleviate diabetes-induced deleterious changes in bone microarchitecture or biomechanical properties.¹³⁶ On the other hand, Bain et al. (1997) reported that oral administration of minocycline (20 mg/day) prevented cancellous bone loss in STZ-induced diabetic rats. The protective effect against diabetes-induced osteopenia was attributed to improved bone formation and mineral apposition.¹³⁷ In line with the histomorphometric findings reported by Bain et al. (1997)¹³⁷, Sasaki and co-authors (1991, 1992) found that orally administering minocycline (20 mg/day) to STZ-induced diabetic rats improved osteoblast morphology and function^138,139 while Kaneko et al. (1990) reported that orally administering minocycline (20 mg/day) did not normalize osteoclast outcomes.¹⁴⁰ Together, these reports demonstrate there is no clear consensus about tetracyclines’ protective effects on diabetes-induced osteopenia.

7. Impact of Tetracyclines on Osteoclastogenesis

Calcium signaling promotes osteoclast differentiation and function. Recognizing that tetracycline antibiotics chelate calcium, early osteoclast researchers evaluated the effect of tetracyclines on cytosolic calcium (Table 5). Donahue et al. (1992) investigated whether minocycline or doxycycline could alter the levels of cytosolic calcium (Ca²⁺) in rat osteoclasts in vitro.¹⁴¹ Neither minocycline nor doxycycline (10 μg/ml) altered the cytosolic Ca²⁺ concentration in osteoclasts. By contrast, Bax et al. (1993) reported that stimulating rat osteoclasts in culture with 10 or 100 mg/ml minocycline increased the concentration of cytosolic Ca²⁺ by approximately 3X fold.²¹ Furthermore, Bax et al. (1993) reported that preincubation with minocycline (1 mg/l), but not doxycycline (1 or 10 mg/l), significantly attenuated the cytosolic Ca²⁺ response in osteoclasts to stimulation with ionic nickel (a surrogate Ca²⁺ receptor agonist). This led the authors to conclude that minocycline attenuates cytosolic Ca²⁺ responses in osteoclasts, which is associated with Ca²⁺ receptor activation.²¹ Zaidi et al. (1993) reported that minocycline (10 mg/l) stimulation exerted morphometric changes in cultured rat osteoclasts, which were comparable to changes elicited by occupancy of the Ca²⁺ receptor.²² Overall, these investigations suggest that tetracycline-calcium interactions can influence osteoclastogenesis.

Table 5:

Tetracycline class antibiotic effects on in vitro osteoclastogenesis

Study	Cell Culture System	Antibiotic	Dose	Treatment Duration	Outcomes
Donahue et al. (1992)141	Primary rat neonatal long bone derived adherent cells	Minocycline or doxycycline	10 μg/ml	500 seconds	No effect on cytosolic calcium
Bax et al. (1993)21	Primary rat neonatal long bone derived adherent cells	Minocycline or doxycycline	0.1, 1, 10, 100 mg/l	1 hour	↑ Cytosolic calcium induced by minocycline
Zaidi et al. (1993)22	Primary rat neonatal long bone derived adherent cells	Minocycline	10 mg/l	80 minutes	Altered cell morphology
Gomes et al. (1984)142	Rat embryonic bone explant cultures treated with 1.0 μg/ml parathyroid hormone	Tetracycline, minocycline, or doxycycline	0.2, 2.0, 20, 200 μg/ml	2 or 5 days	↑ Ca2+ release induced by 20 & 200 μg/ml tetracycline, minocycline, or doxycycline
Zhou et al. (2010)143	Primary rat bone marrow macrophages cultured on bovine bone with 30 ng/ml MCSF, 50 ng/ml RANKL, & 10−7M 1α,25[OH]2D3	Tetracycline	5, 10, 20 μg/ml	Apoptosis: 1 day Formation: 7 days Resorption: 14 days	↑ Osteoclast apoptosis ↓ Osteoclast formation ↓ Bone resorption
Chowdhury et al. (1993)144	Primary chick embryo long bone derived adherent cells cultured on bovine bone	Doxycycline	5 or 15 μg/ml	3, 6, 24 hours	↓ Bone resorption
Bettany et al. (2000)145	Primary murine osteoblasts/osteoclasts co-cultured on ivory wafers	Doxycycline	5 or 15 μg/ml	24 hours	↑ Osteoclast apoptosis ↓ Bone resorption
Holmes et al. (2004)23	Primary human peripheral blood mononuclear cells cultured with 30 ng/ml MCSF & 25 ng/ml RANKL	Doxycycline	2.5 or 20 μg/ml	20 days	↓ Osteoclast formation induced by 2.5 μg/ml ↑ Osteoclast apoptosis induced by 20 μg/ml
Franco et al. (2011)146	Primary murine bone marrow macrophage cells (BMMCs) stimulated with 10 ng/ml MCSF & 50 ng/ml RANKL RAW264.7 murine macrophage cells stimulated with 50 ng/ml RANKL	Tetracycline, doxycycline, or minocycline	0.2 or 2.0 μg/ml	BMMCs: 6 days RAW Cells: 5 days	↓ Osteoclast numbers induced by 0.2 & 2.0 μg/ml doxycycline, 0.2 & 2.0 μg/ml minocycline, or 2.0 μg/ml tetracycline
Nagasawa et al. (2011)24	Primary murine bone marrow cells stimulated with 10 nM 1α,25[OH]2D3 or 20 ng/ml RANKL	Tetracycline, doxycycline, or minocycline	2.5, 5, 10 μM	Formation: 7 days Viability: 24 hours	↓ Osteoclast formation induced by 2.5, 5, & 10 μM tetracycline; 2.5, 5, & 10 μM doxycycline; or 5 & 10 μM minocycline No effect on cell viability
Kim et al. (2019)25	Primary murine bone marrow macrophage cells stimulated with 30 ng/ml MCSF & 100 ng/ml RANKL	Minocycline or tigecycline	1, 2.5, 5 μM	3 days	↓ Osteoclast formation induced by 1, 2.5, & 5 μM minocycline or 1, 2.5, & 5 μM tigecycline

Studies have also shown that tetracycline and its derivatives can have negative effects on osteoclast formation and resorptive function (Table 5). Gomes et al. (1984) assessed the effects of tetracycline, minocycline, and doxycycline (0.2–200 μg/ml) on bone resorption in parathyroid hormone-induced bone organ cultures. Bone resorption was measured by radiolabeled calcium release from embryonic bones. Tetracycline, minocycline and doxycycline had no effect at lower concentrations (0.2 or 2.0 μg/ml), but inhibited bone resorption at higher concentrations (20 or 200 μg/ml).¹⁴² Zhou et al. (2010) investigated the effects of tetracycline on osteoclastogenesis in rat bone marrow macrophages stimulated with MCSF and RANKL, which are critical factors for inducing in vitro osteoclastogenesis. Tetracycline at 5 μg/ml did not influence osteoclast formation, resorption, or apoptosis. However, tetracycline inhibited osteoclast formation, attenuated resorptive capacity, and induced apoptosis at concentrations from 10–20 μg/ml.¹⁴³ Chowdhury et al. (1993) cultured osteoclasts isolated from chick embryo long bones on devitalized bovine cortical bone slices to evaluate the impact of doxycycline on osteoclast function. The addition of doxycycline to the culture system reduced the osteoclast resorptive capacity. Five μg/ml doxycycline reduced the number of resorptive pits by about 50%, and 15 μg/ml doxycycline largely prevented pit formation.¹⁴⁴ By co-culturing murine osteoblasts / osteoclasts on ivory wafers, Bettany et al. (2000) similarly demonstrated that doxycycline inhibits osteoclast resorptive activity in vitro. Five μg/ml doxycycline decreased the area resorbed by roughly 50%, and 15 μg/ml doxycycline prevented bone resorption altogether.¹⁴⁵ Holmes et al. (2004) investigated doxycycline effects in human peripheral blood mononuclear cell-derived osteoclasts. Mononuclear cells were isolated from the peripheral blood and stimulated with MCSF and RANKL. Doxycycline had no effect on osteoclast differentiation at concentrations less than 250 ng/ml, but reduced osteoclast formation by 50% at 2.5 μg/ml. Doxycycline treatment caused apoptosis in cultured human osteoclasts when concentrations exceeded 20 μg/ml.²³ These in vitro studies support the notion that tetracyclines have anti-osteoclastogenic effects.

Subsequent studies investigated the mechanism by which tetracyclines attenuate osteoclastogenesis (Table 5). Franco et al. (2011) evaluated the effects of tetracycline, minocycline, and doxycycline on RANKL-induced RAW264.7 cell osteoclast cultures. Osteoclast numbers per well were reduced by 0.2 and 2.0 μg/ml minocycline, 0.2 and 2.0 μg/ml doxycycline, and 2.0 μg/ml tetracycline.¹⁴⁶ The authors found that doxycycline downregulated RANKL-induced osteoclastogenesis through inhibitory effects on RANKL-induced MMP-9 enzyme activity. The inhibitory effects were independent of the transcription factor, NFATc1, which is critical for osteoclast differentiation.¹⁴⁶ Nagasawa et al. (2011) evaluated the effects of tetracycline, doxycycline, and minocycline on murine bone marrow cell cultures. Osteoclastogenesis was induced in the culture system via 1α,25[OH]₂D₃ or RANKL. Tetracycline drug concentrations at 2.5 / 5 / 10 μM inhibited osteoclast formation in a dose-dependent manner. The authors concluded that minocycline inhibits RANKL-induced osteoclastogenesis via the downregulation of Nfatc1.²⁴ Kim et al. (2019) reported similar outcomes in murine bone marrow macrophage cultures stimulated with MCSF and RANKL. Minocycline or tigecycline (1 / 2.5 / 5 μM) treatment abrogated osteoclast formation through suppression of NFATc1 signaling. The authors speculated that minocycline and tigecycline inhibit osteoclast differentiation by suppressing MMP-9-mediated histone H3 cleavage.²⁵ These reports provide evidence that NFATc1 and MMP-9 may contribute to tetracycline inhibitory actions on osteoclastogenesis.

8. Impact of Tetracyclines on Osteoblastogenesis

The effects of tetracycline antibiotics on osteoblasts (Table 6) have received far less attention than osteoclasts (Table 5). Duewelhenke et al. (2007) found that incubating primary human osteoblasts with tetracycline (60–80 μg/ml) inhibited cell proliferation by 20%.²⁶ Gomes et al. (2007) investigated the effect of doxycycline (1–25 μg/ml) and minocycline (1–50 μg/ml) on the proliferation, differentiation, and function of human bone marrow-derived osteoblastic cells in culture. A 1 μg/ml dose of doxycycline or minocycline increased the proliferation of osteoblastic cells without altering their functional activity. However, higher doses (≥5 μg/ml) inhibited osteoblast function.²⁷ Park et al. (2011) assessed the effects of tetracycline, minocycline, and doxycycline (10 or 100 μM) on MC3T3-E1 murine osteoprecursor cells. Tetracycline at 10 or 100 μM, doxycycline at 100 μM, and minocycline at 100 uM reduced cell viability and mineralization.²⁸ Rathbone et al. (2011) similarly showed that high concentrations of doxycycline and minocycline suppress human osteoblast cell numbers and ALPase activity in vitro.²⁹ Kim et al. (2019) studied minocycline and tigecycline effects on primary murine calvaria derived osteoblasts. Whereas minocycline (2.5 or 5 μM) had no effect, tigecycline (2.5 or 5 μM) attenuated osteoblast function.²⁵ Interestingly, these investigations indicate that tetracycline antibiotics can have divergent dose-dependent effects on osteoblastogenesis.

Table 6:

Tetracycline class antibiotic effects on in vitro osteoblastogenesis

Study	Cell Culture System	Antibiotic	Dose	Treatment Duration	Outcomes
Duewelhenke et al. (2007)26	Primary human femur derived mesenchymal stem cells stimulated with 100 nM dexamethasone, 50 uM ascorbic acid, & 10 mM β-glycerophosphate	Tetracycline	60–80 μg/ml	48 hours	↓ Osteoblast proliferation
Gomes et al. (2007)27	Primary human bone marrow derived osteoblastic cells stimulated with 50 mg/ml ascorbic acid, 10mM β-glycerophosphate, & 10 nM dexamethasone	Minocycline or doxycycline	Minocycline: 1, 5, 10, 25, 50 μg/ml Doxycycline: 1, 5, 10, 25 μg/ml	35 days	↑ Osteoblast proliferation induced by 1 & 5 μg/ml doxycycline or 1, 5, & 10 μg/ml minocycline ↓ Osteoblast proliferation induced by 10 & 25 μg/ml doxycycline or 25 & 50 μg/ml minocycline ↓ Osteoblast function induced by 25 μg/ml doxycycline or 50 μg/ml minocycline
Park et al. (2011)28	MC3T3-E1 murine osteoprecursor cells stimulated with 50 μg/mL ascorbic acid & 10 mM β-glycerophosphate	Tetracycline, minocycline, or doxycycline	10 or 100 μM	4 days	↓ Osteoblast viability and mineralization induced by 10 & 100 μM tetracycline, 100 μM minocycline, or 100 μM doxycycline
Rathbone et al. (2011)29	Human osteoblasts cultured with 50 μg/ml ascorbic acid, 5 mM glycerophosphate, & 10 nM dexamethasone	Minocycline or doxycycline	Minocycline: 10, 100, 200, 500 μg/ml Doxycycline: 10, 100, 200 μg/ml	10 or 14 days	↓ Osteoblast numbers and function induced by 10 – 500 μg/ml minocycline or 10 – 200 μg/ml doxycycline
Kim et al. (2019)25	Primary murine calvaria derived osteoblasts cultured in 50 μg/mL ascorbic acid & 10 mM β-glycerophosphate	Minocycline or tigecycline	2.5 or 5 μM	14 or 19 days	Minocycline did not alter osteoblast function ↓ Osteoblast function induced by 2.5 & 5 μM tigecycline

9. Future Directions

9.1. Skeletal retention of tetracyclines

Tetracyclines have a high affinity for the bone matrix and are retained within the skeleton for extended periods after discontinuing treatment^10,11. This raises the question whether tetracyclines retained in the bone matrix can have prolonged effects on bone modeling and remodeling processes.^72,73 Recognizing that long durations of therapy results in cumulative retention in the skeleton,^10,11 tetracyclines administered for acne are likely retained in the bone matrix at high concentrations. Roughly 10% of the adult human skeleton is remodeled each year, which includes about 30% of trabecular bone and 3% of cortical bone.^147–149 Given that tetracyclines are prominently deposited on trabecular surfaces^8,9 due to the higher remodeling rate of cancellous bone,^147,149 this highlights the potential for tetracycline drugs released from the bone matrix to have lasting effects. Investigations are needed to understand mechanisms by which tetracycline drugs bind the bone matrix, the concentration of drug retained in the skeleton, and how drug / metabolite release from the bone impacts bone modeling / remodeling processes.

9.2. Skeletal maturation and homeostasis

The impact of tetracyclines on bone mass accrual and skeletal maturation during the adolescent and young adult phase is currently unknown. Considering that roughly 40% of peak bone mass is accrued during adolescence and young adulthood,^150–153 the administration of tetracycline antibiotics during this developmental period could dysregulate skeletal growth and maturation. The effects of tetracycline antibiotics on bone remodeling and homeostasis are currently unclear in the mature adult skeleton. Appreciating that tetracycline drugs are administered for extended durations as a treatment for rosacea in adults, there may be detrimental consequences for bone remodeling and the maintenance of bone mass in the healthy adult skeleton.

9.3. Signaling at bone cells

The mechanisms by which tetracyclines affect bone cells are not well understood. Early investigators studying minocycline-induced changes in the levels of cytosolic Ca²⁺ in osteoclast cultures speculated that tetracyclines have a partial agonistic effect on the osteoclast Ca²⁺ receptor.^21,22 Minocycline and doxycycline have been reported to inhibit RANKL-induced osteoclastogenesis by blocking downstream MMP-9 enzyme activity, but it is unclear whether this inhibition occurs through the NFATc1 signaling pathway.^24,25,146 The signaling mechanisms by which tetracyclines influence osteoblastogenesis have not been investigated. The impact of tetracyclines on osteocytes, which account for 90% of bone cells, also has not been explored. Considering that osteocytes express biologic factors that critically regulate osteoblast and osteoclast activity,^154,155 future investigations need to delineate the impact of tetracyclines on osteocyte biology and coupled osteoblast-osteoclast bone remodeling processes.

In vitro investigations have shown that tetracyclines have direct effects on osteoclasts and osteoblasts, but many of these study outcomes were dependent on supraphysiologic drug concentrations (Table 5 and 6). Clinical pharmacokinetic data indicates that long-term administration of doxycycline or minocycline (100 mg every 12 hours) results in steady-state serum concentrations ranging from 1.0 to 3.5 μg/ml.¹⁵⁶ Future in vitro bone cell investigations should utilize tetracycline concentrations within the range detected in human serum for study findings to be clinically relevant.

Clinical pharmacokinetic studies have shown that tetracycline drugs are broken down into metabolites, which may have implications for bone cell signaling. There is one known metabolite of tetracycline that has been found in humans, Δ-epitetracycline.⁵² Whereas there are no known metabolites of doxycycline, six metabolites of minocycline have been detected in humans. The three major minocycline metabolites are 9-Hydroxyminocycline, N7-Demethylminocycline, and N4-Demethylminocycline.^52,156–158 In order to advance knowledge about the signaling of tetracyclines at bone cells, future studies must begin to investigate how tetracycline drug metabolites impact bone cell physiology.

9.4. Gut microbiota and bone

The link between gut microbiota and bone has recently been recognized. Investigations comparing germ-free mice to conventional mice have shown that immunomodulatory actions imparted by the gut microbiota regulate the development and homeostasis of skeletal tissues^159–163. Disruption of the gut microbiota by antibiotics has been shown to dysregulate osteoimmune crosstalk in post-pubertal skeletal maturation. A broad-spectrum antibiotic cocktail (vancomycin, imipenem/cilastatin, and neomycin) administered to sex-matched C57BL/6T mice from age 6 to 12 weeks induced phylum shifts in the gut bacteriome and led to suppression of bone mass accrual in the late growing skeleton¹⁶⁴. This work indicates that antibiotic-induced changes in the gut microbiota can have indirect effects on the skeleton. Research is needed to determine whether tetracycline drug-induced perturbations in the gut microbiota are linked to alterations in skeletal maturation and homeostasis.

9.5. Third generation tetracycline antibiotic effects on bone

The effects of newer generation tetracycline antibiotics on the skeleton are unclear. Tigecycline is typically administered for infections resistant to earlier generation tetracycline antibiotics.^45,46 Several in vitro investigations have begun to evaluate the efficacy of tigecycline for the treatment of osteomyelitis,^165–167 but the effects of the drug on normal bone physiology is poorly understood. Sarecycline was recently FDA approved for the treatment of moderate-to-severe acne in patients older than nine years of age.⁴¹ Considering that sarecycline is administered for 3 to 12 months for the treatment of acne in adolescents and young adults, there could be implications for late skeletal growth and maturation. There are currently no known studies that have evaluated the impact of sarecycline, omadacycline, or eravacycline on bone cells or skeletal metabolism.

10. Conclusion

The mechanisms by which tetracycline drugs bind and are released from the bone matrix is not well understood. The effects of tetracycline, minocycline and doxycycline on bone cell signaling and their impact on physiologic skeletal maturation and homeostasis are unclear. The role of the gut microbiota in tetracycline antibiotic effects on bone metabolism has yet to be investigated.

Newer third generation tetracycline antibiotics (i.e., sarecycline, omadacycline, eravacycline) are being utilized for clinical care, but their effects on bone have not been studied. Future research is needed to elucidate the molecular underpinnings of tetracycline drug actions on bone cells and their impact on skeletal physiology. Tetracyclines’ affinity for and retention in the bone matrix underscores that these drugs may have lasting effects on bone metabolism.

Highlights.

• Tetracyclines retained in bone may have lasting effects on skeletal metabolism.

• Tetracyclines’ impact on skeletal maturation and homeostasis is poorly understood.

• Mechanisms regulating tetracyclines’ actions on bone cells are largely unknown.

• The gut microbiota’s role in tetracycline antibiotic actions on bone is undefined.

Abbreviations

MCSF

Macrophage Colony-Stimulating Factor

RANKL

Receptor Activator of Nuclear Factor κβ Ligand

OVX

Ovariectomized

STZ

Streptozotocin