Sensitive Detection of β-Carotene in Chromogenic and Non-Chromogenic Mycobacteria by HPLC-DAD and UHPLC-MS

Abstract

Carotenoid pigments are widely distributed in nature and play a crucial role in protecting organisms from photodynamic damage. However, the characterization of carotenoid production in clinically relevant mycobacteria has been limited due to the low sensitivity of conventional detection methods. We present a descriptive analysis of carotenoid production in seven mycobacterial isolates from the scotochromogenic, photochromogenic, and non-chromogenic groups. To achieve this, we used a combination of High-performance liquid chromatography with diode-array detection (HPLC-DAD) and Ultra-high performance liquid chromatography–mass spectrometry (UHPLC-MS) to detect carotenoids pigments. Mycobacterium tuberculosis (MTB) and Mycobacterium bovis (MB) (non-chromogenic mycobacteria) produced β-carotene when cultured in the absence of light, at levels comparable to those of photochromogenic mycobacteria such as M. marinum (MM) and M. kansasii (MK). The highest levels of carotenoids were found in scotochromogenic species M. avium (MAV) and M. gordonae (MGOR). Conversely, M. abscessus (MABS), a non-chromogenic species in which no β-carotene was detected, served as a negative control for matrix effects. As expected, the use of highly sensitive analytical techniques such as HPLC-DAD and UHPLC-MS significantly enhanced the detection of β-carotene compared to visual pigment assessment. These methods allowed the detection of basal β-carotene levels even in mycobacteria classified as non-chromogenic. The proposed analytical approach provides a robust research tool to understand the effects of different stimulus that may alter the cell physiology in terms of pigment production.

1. Introduction

Carotenoids are one of the most important groups of natural pigments due to their wide distribution, structural diversity, and numerous functions [1]. β-carotene, a pigment within this group, is a tetraterpene that is biochemically synthesized from eight isoprene units. These units are cyclized to form beta rings at both ends of the molecule and contain a linear structure of C40 hydrocarbons with three to fifteen conjugated double bonds [2], (Figure 1). The number of double bonds determines the spectral properties of the carotenoids, allowing them to absorb light at wavelengths between 400 and 500 nm, which confers a characteristic color to the colony or microorganisms that synthesize them (Figure 2) [3]. Regarding their functions, carotenes are involved in numerous biochemical processes, such as subcellular orientation and regulation, free radical scavenging and cell protection against photodynamic lesions and pH changes [4,5]. The production of these pigments is therefore an important characteristic of several microorganisms facing these challenges such as yeasts, filamentous fungi, micro-algae and bacteria [6,7].

In the genus Mycobacterium, the presence of carotenes has been widely described [2,3,8,9]. Beyond their phenotypic relevance, carotenoid biosynthesis in mycobacteria is genetically encoded by a conserved crt gene cluster, comprising the enzymes crtE (geranylgeranyl diphosphate synthase), crtB (phytoene synthase), crtI (phytoene desaturase), and crtY (lycopene cyclase), which collectively mediate the sequential conversion of isoprenoid intermediates derived from the DOXP pathway into β-carotene [3]. Comparative analyses of Mycobacterium marinum (MM), M. kansasii (MK), and M. vaccae (MV) have demonstrated that these genes are organized in a similar genomic arrangement, supporting the notion that carotenoid production is a stable and conserved trait among pigment-producing mycobacteria. However, as noted by Robledo et al. [3], some species within the genus lack visible pigmentation, suggesting that while the genetic machinery for carotenoid synthesis may persist, its transcriptional activation or enzymatic efficiency differs across species. This pattern indicates that evolutionary conservation of the crt cluster is coupled with regulatory diversification, allowing photochromogenic species such as MM and MK to produce β-carotene under light stimulation, whereas nonchromogenic species like M. tuberculosis (MTB) and M. bovis (MB) likely retain these genes for non-photogenic functions related to oxidative stress protection. This genetic heterogeneity provides a molecular explanation for the phenotypic variation described by Runyon, thereby linking genotypic regulation with pigment expression in the genus Mycobacterium.

Over the last five decades, the Runyon system has therefore been used to classify mycobacteria based on visually detected pigment production (Figure 2) into the following categories: (i) scotochromogenic, those whose production is independent of light stimulation; (ii) photochromogenic, those that produce pigmentation in response to exposure to light; (iii) those without visually detectable pigment (non-chromogenic) regardless of light stimulation. This classification system was complemented by the analysis of the proliferation rate: slow-growing (taking more than a week to develop colonies) or fast-growing (taking less than a week to develop colonies) [10]. These characteristics, together with a series of biochemical tests, acquired taxonomic value and contributed to species identification. Despite still being useful today, this method is unsatisfactory due to its basis on visual detection, which lacks sensitivity and reproducibility in determining the presence or absence of pigment. Thus, thin layer chromatography (TLC) was incorporated to detect carotenoids in different mycobacterial species [11,12]. Although this resulted in a significant improvement, TLC only allowed qualitative determinations, or, in the best situation, semi-quantitative determinations of carotenes, with poor sensitivity.

High-performance liquid chromatography HPLC-based analyses of different types of crude samples for the detection of carotenoids are important in several applications, such as food chemistry and medicine [13,14,15]. This method offers significant advantages in terms of simplicity, speed, cost, sensitivity, specificity, precision, and sample preservation. The use of a diode array detector (DAD) coupled to HPLC enables continuous data collection, which helps determine peak purity. Liquid chromatography–mass spectrometry (LC-MS) is even more specific and sensitive, providing information on carotenoid molecular mass and fragmentation patterns, with a detection threshold in the range of 500 fmol for individual carotenoids [15].

Based on the need of more sensitive and reproducible methods to detect β-carotene for identifying mycobacterial species pathogenic to humans, we developed an optimized pipeline for its detection. The method was validated for detecting basal levels of β-carotene production in seven different species: MTB, MB, MM, MK, M. avium (MAV), M. gordonae (MGOR) and M. abscessus (MABS).

2. Materials and Methods

2.1. Mycobacterial Material

Seven representative species of mycobacteria were used in this study. Three different isolates were obtained from clinical samples, which were stored at −70 °C in the strain bank of the Bacteriology and Mycobacteria Unit of the Corporation for Biological Research (CIB) in Medellín, Colombia. The species of clinical interest selected were MTB, MB, MM, MK, MAV, MGOR and MABS (

Table A1. Clinically relevant mycobacterial species and isolates *.

Species	Isolates
Mycobacterium tuberculosis (MTB)	MTB469
	MTB509
	MTB204
Mycobacterium bovis (MB)	MB33
	MB34
	MB35
Mycobacterium marinum (MM)	MM221
	MM577
	MM1004
Mycobacterium kansasii (MK)	MK808
	MK1023
	MK1030
Mycobacterium avium (MAV)	MAV431
	MAV437
	MAV570
Mycobacterium gordonae (MGOR)	MGOR1166
	MGOR1168
	MGOR1169
Mycobacterium abscessus (MABS)	MABS1158
	MABS1124
	MABS1064

* The isolate codes correspond to their identification in the CIB strain collection.

).

2.2. Sample Preparation

Each isolate from the strain bank was reconstituted in Middlebrook 7H9 medium for its functional and metabolic reactivation over a period ranging from one to two weeks, depending on the mycobacterial species. Then, a scale culture was started, for which an initial inoculum of 1 mL of bacterial cells was placed in 50 mL of medium until it reached a McFarland turbidity standard of 1, equivalent to 3 × 10⁸ CFU/mL, and subsequently transferred to 500 mL of medium. All the cultures in each of their stages were kept in conditions of total darkness, constant shaking (85 RPM) and incubation at 37 °C, until reaching a stationary growth phase, considering two weeks for the fast-growing species and four weeks for the slow-growing species. For β-carotene extraction, 3 × 10⁸ CFU/mL in the stationary phase were harvested by centrifugation (Sorvall legend RT refrigerated, Boston, MA, USA) at 4150 RPM × 15 min at 4 °C, and the cell pellet was separated from the supernatant. The cell pellet was homogenized with 1 mL of sterile isotonic saline solution (NaCl 0.85%) and 200 µL of lysozyme was added at 20 mg/mL (Sigma, Darmstadt, Germany; L-6876), followed by incubation at 37 °C overnight in the dark. β-carotene extraction was based on a standard method (13), which involves saponification of the cell pellet with 100 mL of absolute ethanol (Cat. 107017, Merck, Darmstadt, Germany;) and 20 mL of potassium hydroxide (KOH) (Cat. 105033, Merck, Germany) at 60%, mixed with 1 g of ascorbic acid (Cat. 47863, Sigma, Germany) to prevent oxidation during the extraction process. The mixture was incubated at 30 °C for 1 h, then passed 3 times through a separatory funnel with 50 mL volumes of hexane until the solution was decolorized. The pigmented fraction corresponding to the hexane phase, was recovered each time. The combined hexane phases were collected in an amber flask and evaporated under reduced pressure at 40 °C and 40 RPM in a rotary evaporator (BUCHI, Flawil, Switzerland; R-210) until complete.

2.3. HPLC-DAD Analysis

The analysis was performed on an Agilent 1200 series chromatograph (Germany) with a C30 column (5 μm; 150 × 4.6 mm Thermo Scientific, Waltham, MA, USA) and equipped with a diode-array detector. An isocratic method 75:24.9:0.1 ethanol, hexane and 2-propanol, respectively, over 8 min at a flow rate of 1 mL min⁻¹ at 20 °C was used as the mobile phase. Detection was performed at 450 nm. For method validation, the following aspects were considered: linearity, limits of detection and quantification, precision and accuracy, percentage of recovery and stability. In this study, particular emphasis was placed on the linearity parameter because it ensures the reliability and reproducibility of quantification across the tested concentration range. Establishing linearity was critical to confirm that the detector response was directly proportional to β-carotene concentration, which in turn allowed accurate comparisons of pigment levels among different Mycobacterium species. This parameter also minimized potential bias from matrix effects and provided the necessary robustness for subsequent biological interpretations. As a reference, the commercial β-carotene (type II, synthetic, ≥95%, C4582, Sigma-Aldrich, Steinheim, Germany) was used to prepare the standard curve for quantification. For method calibration, a standard curve with a scale that moved from 0.0098 µg to 10 µg of β-carotene was prepared by mobile phase dilution from the standard solution. HPLC methods use either isocratic or gradient mobile phases in reversed-phase or normal-phase mode [13,16]. Since carotenoids are easily oxidized, it is useful to add antioxidants to the extraction solvent and mobile phase and to keep a low, constant column temperature [17].

2.4. UHPLC-MS Analysis

To unambiguously confirm the identity of the pigment, mass spectra of carotenoids were recorded using a Shimadzu UHPLC with an LCMS 2020 mass detector. The DL temperature was 275 °C, and the heat block was set to 425 °C, with a nebulizing gas flow of 1.5 L/min. Mass spectra were recorded over a range of 200 to 600. Chromatographic separation was achieved using an XDB-C18 column (1.8 μm; 2.1 × 50 mm ZORBAX Eclipse) at 40 °C with an isocratic method 55:45 acetonitrile and methanol, respectively, over 10 min at a flow rate of 0.5 mL min⁻¹.

A β-carotene standard was used to determine the peak retention times. HPLC column properties were checked at regular intervals and re-evaluated using the standard. Pigment peak identification was performed using HPLC-DAD with the calibration curve of the standard, allowing for sample quantification and confirming the presence of β-carotene via UHPLC-MS.

3. Results

3.1. Analytical Characteristics of the Beta-Carotene Detection Method by HPLC-DAD

For the linearity parameter, calibration curves were prepared in triplicate within the defined range. All curves demonstrated excellent linearity (R² > 0.999) across the analyzed range, from 0.0098 µg/mL to 10 µg/mL (

Table A4. Standard curve calculations using commercial β-carotene.

Concentration (μg/mL)	Average Areas	Experimental Mean Concentration
0.0098	3.547	0.0096
0.0195	6.850	0.0199
0.039	12.447	0.0371
0.078	25.840	0.0789
0.156	51.883	0.1606
0.312	108.937	0.3368
0.624	203.580	0.6300
1.248	402.617	1.2504
2.496	811.213	2.5284
4.992	1570.057	4.8936
9.984	3226.123	10.0449

, Figure A1). The limit of detection (LOD) was determined as 0.003 µg/mL (signal area of 1584 AU), and the limit of quantification (LOQ) as 0.01 µg/mL (signal area of 3168 AU). Both values fall below the lowest point of the calibration range, confirming the sensitivity of the method. In Mycobacterium, there are no previous references for the minimum detectable levels of β-carotene; therefore, the LOQ obtained here demonstrates that the proposed method is suitable for detecting β-carotene without matrix interference. It is important to note that previously reported LOD/LOQ values for β-carotene from other biological sources are expressed in terms of concentration (0.157 µg/mL for LOD and 0.52 µg/mL for LOQ) [18].

Precision and accuracy were assessed in triplicate, considering both repeatability and reproducibility [%Re]) and evaluated using the coefficient of variation and relative error percentages, which were less than 20% (according to the Bioanalytical Method Validation, FDA. 2018), corresponding to the maximum accepted variation (

Table A5. Inter-assay precision and accuracy (n = 9, 3 trials).

Theoretical Concentration (µg/mL)	Determined Concentration (µg/mL)
0.1563	Mean=	0.17
	%CV=	9.0
	%ER=	11.9
2.5	Mean=	2.35
	%CV=	9.6
	%ER=	5.8
10.0	Mean=	10.90
	%CV=	3.3
	%ER=	9.0

, Figure A2). Three concentrations were used (0.039 µg/mL, 0.625 µg/mL and 10 µg/mL of the standard) to determine the %Re, and we worked with standards in a diluent solution as a reference value, which was equivalent to 100%. The average %Re of the applied method was 98.6% with a coefficient of variation of 11.4%. The matrix for this case did not show appreciable interferences, as evidenced by this result (≈100%). The value of 100% indicates that no measurable matrix interference was observed, and that the recovery rate was equivalent to that obtained with the standard alone. Thus, the term reflects both the recovery efficiency and the absence of matrix effects in our quantification.

3.2. Quantification of B-Carotene by HPLC-DAD

Among the 21 isolates evaluated from seven species of mycobacteria, two isolates from two species evaluated showed high β-carotene content, around 4 µg/mL. Figure 3 shows the β-carotene content per species. MGOR exhibited the highest detected amounts (Figure A3), while MABS was used as the control since no pigment was detected in the evaluated isolates. Importantly, basal levels of β-carotene were also observed in the photochromogenic species MM and MK, even under dark growth conditions.

Despite detecting pigments in six of the seven evaluated species, intraspecies analysis revealed significant differences between isolates. Figure 4 shows the detection of β-carotene for each isolate within each species. For MTB and MB, one isolate from each species exhibited either no or low β-carotene detection (Figure A4). However, the other two examined isolates showed pigment levels comparable to the average range found among species classified as chromogenic or photochromogenic (Figure A5), corresponding to 0.7–1.3 µg/mL. Calculations were based on a standard curve generated using commercial β-carotene (Table A4, Figure A1).

3.3. Detection of β-Carotene Confirmed by UHPLC-MS

Samples quantified by HPLC-DAD were analyzed using the UHPLC-MS method to confirm whether the detected pigment corresponded to β-carotene.

Table 1. Detection levels by HPLC-DAD of β-carotene and confirmation by UHPLC-MS in the 3 isolates of the mycobacteria of clinical interest.

Species/Isolates	Area	µg/mL	SD	CV	Mean	UHPLC-MS
Non-chromogens
MTB469R1	227.3	0.71	0.03	4.35	0.70	+
MTB469R2	233.24	0.72
MTB469R3	214.15	0.66
MTB509R1	17.66	0.09	0.01	12.50	0.08	+
MTB509R2	16.3	0.07
MTB509R3	15.48	0.08
MTB240R1	338.7	1.05	0.09	8.15	1.13	+
MTB240R2	395.83	1.23
MTB240R3	354.28	1.10
MB33R1	<1.59	<0.003	0.00			−
MB33R2	<1.59	<0.003
MB33R3	<1.59	<0.003
MB34R1	224.7	0.70	0.17	23.53	0.71	+
MB34R2	286.55	0.89
MB34R3	178.93	0.55
MB35R1	380.28	1.18	0.08	6.66	1.23	+
MB35R2	383.57	1.19
MB35R3	427.58	1.33
Scotochromogens
MGOR1168R1	649.63	2.02	0.16	8.3	1.91	+
MGOR1168R2	556.68	1.73
MGOR1168R3	640.64	1.99
MGOR1169R1	532.13	1.65	0.14	8.4	1.65	+
MGOR1169R2	576.92	1.79
MGOR1169R3	487.34	1.51
MGOR1166R1	1594.69	4.96	0.95	22.6	4.21	+
MGOR1166R2	1010.29	3.14
MGOR1166R3	1460.41	4.54
Photo-chromogens
MM221R1	256.3	0.80	0.07	9.17	0.72	+
MM221R2	217.8	0.68
MM221R3	221.6	0.69
MM577R1	54.88	0.17	0.01	7.00	0.16	+
MM577R2	53.7	0.17
MM577R3	48.1	0.15
MM1004R1	230.93	0.72	0.05	6.50	0.72	+
MM1004R2	215.33	0.67
MM1004R3	245.21	0.76
MK808R1	343.7	1.07	0.16	13.28	1.22	+
MK808R2	384.64	1.19
MK808R3	446.9	1.39
MK1023R1	98.7	0.31	0.04	10.54	0.35	+
MK1023R2	120.98	0.37
MK1023R3	116.33	0.36
MK1030R1	343.25	1.07	0.09	8.32	1.08	+
MK1030R2	319.29	0.99
MK1030R3	376.59	1.17
MAV431R1	379.43	1.18	0.12	9.11	1.28	+
MAV431R2	403.9	1.25
MAV431R3	453.1	1.41
MAV437R1	517	1.61	0.13	7.45	1.72	+
MAV437R2	598.7	1.86
MAV437R3	547.73	1.70
MAV570R1	314.34	0.98	0.15	18.11	0.83	+
MAV570R2	218.1	0.68
MAV570R3	266.22	0.83

MM (M. marinum); MTB (M. tuberculosis); MB (M. bovis); MK (M. kansasii); MAV (M. avium); MGOR (M. gordonae). SD: Standard deviation; CV: coefficient of variation (percentage).

presents both the quantified numerical values and qualitative data regarding the presence or absence of pigment in each isolate of each species. The analytical characteristics of the UHPLC-MS β-carotene detection method include general parameters (

Table A2. Chromatographic conditions and the parameters of the UHPLC-MS detector.

Parameters	Characteristics
Mobile phase	Acetonitrile: Methanol 55:45
Flux	0.5 mL/min
Injection Volume	5 uL
Column	ZORBAX Eclipse XDB-C18 2.1 × 50 mm 1.8 Micron
Column Temperature	40 °C
Retention time	8 min
Run Time	10 min
Mass detector parameters:
Interface
ESI
DL temperature	275 °C
Nebulizing gas flow	1.5 L/min
Heat block	425 °C
Drying gas flow	10 L/min
Event 1 SIM (m/z)	536.35
Event 2 Scan (m/z)	200–600

), system suitability (

Table A3. Parameters and acceptance criteria for system suitability.

Parameter	Acceptance Requirements
Mass accuracy for the standard	535.85–536.85 m/z
Signal/noise ratio (S/N)	≥2.0
Percentage of error in final retention time	≤10%

), selectivity (Appendix A.4), and detection limits (Appendix A.5).

4. Discussion

To assess the ability of the proposed HPLC-DAD method to detect β-carotene in clinically relevant samples, analytes were initially extracted from cultures of three different clinical isolates per species, each tested in triplicate. MGOR, a primarily environmental and opportunistic scotochromogenic species, exhibited the highest β-carotene levels (3 µg/mL) among the tested species, corresponding to the amount produced through constitutive biosynthesis. For the photochromic species MM and MK, the obtained values were 0.7 µg/mL and 1.22 µg/mL, respectively. These lower values, compared to those of MGOR, are comparable to those of non-chromogenic species. In addition, this is consistent with the fact that the extracts were not exposed to light during their preparation. Remarkably, since the 1970s, the presence of β-carotene has been documented in both species, but there is no reference to quantified detectable levels [12,19].

Gao et al. [20] showed that the absence of pigment increases cellular susceptibility to singlet oxygen and identified two homologous (putative) genes within an operon in MTB, which may be associated with pigment synthesis. These results are consistent with the baseline β-carotene levels observed in both MTB and MB (members of the MTB complex), which are comparable to those found in MM (Figure 3). However, some variability was observed among MTB and MB isolates, despite their being cultured under identical amplification and growth conditions. This variability could reflect differences among the original clinical sources, as mycobacterial isolates often display phenotypic diversity influenced by their genetic backgrounds, infection histories, and adaptive responses to host environments. Nonetheless, these interpretations remain speculative in the absence of genotypic data. To determine whether such heterogeneity arises from true genetic or regulatory differences, future studies should incorporate whole-genome sequencing or transcriptomic profiling to identify potential determinants underlying pigment variability.

Regarding MAV, it is one of the most important NTMs causing mycobacteriosis and is classified as a non-chromogenic species. However, both pigmented and non-pigmented colonies have been observed in isolates of this species and have been associated with antibiotic resistance [21]. At the basal level, MAV produces significant amounts of β-carotene compared to the species included in this analysis, linking it to scotochromogenic species with constitutive pigment production (Table 1). For MABS, β-carotene was not detected at the basal level within the range and conditions established for this method, which is consistent with its classification as a non-chromogenic species. However, a study by Saviola & Felton [22] reported that MABS presents pigmentation under acidic pH conditions.

These inter-species comparisons provide a baseline for understanding β-carotene production across clinically and environmentally relevant mycobacteria. Building upon this, a closer examination of intra-species variability revealed additional layers of complexity. Notably, significant differences were observed among clinical isolates belonging to the same species (Figure 4). For instance, within the MB group, the coefficient of variation in β-carotene concentrations reached 23.53%, reflecting substantial inconsistency among replicates of MB34, while no analyte was detected in MB33 using the applied method.

Although most species presented acceptable coefficients of variation (<20%), MAV 570 approached this upper threshold (18.11%), further reinforcing the presence of intra-species heterogeneity (Table 1). This challenges the assumption that pigment production is homogeneously expressed within a species, particularly for members of the MTB complex. Such findings suggest a strain-specific metabolic plasticity that may reflect differences in gene regulation, mutations in biosynthetic pathways, or environmental adaptation.

Importantly, the presence or absence of β-carotene as detected by HPLC-DAD was confirmed using UHPLC-MS, which provided rapid, high-specificity spectra for the target compound (m/z 536). This cross-validation confirmed that the detected signals corresponded to β-carotene and not to other structurally similar molecules, reinforcing the biological relevance of the findings.

The clinical isolates used in this research were obtained from the CIB collection of microorganisms, in which they are stored at −70 °C since the time of isolation and identification, without being exposed to successive passages. This preservation method minimizes alterations in phenotypic expression, allowing the retention of pigments anchored in the cell membrane as a result of mechanisms activated prior to cryopreservation. The mechanisms inducing pigment synthesis in these isolates remain unknown, including potential factors such as exposure to UV light [18], pH changes [9,22] and intracellular response in macrophages [20].

While pigmentation remains a fundamental taxonomic criterion in the classification of mycobacteria, our findings reinforce that visual inspection alone is insufficient to confirm pigment production. The application of HPLC-DAD in this study enabled the detection of basal β-carotene levels in species considered non-pigmented, such as MTB and MB. In carotenoid analysis, saponification is commonly used to release esterified forms and reduce the risk of artifact formation. However, several studies have shown that saponification can adversely affect analytical precision [23]. Moreover, because carotenoids are often bound to fatty acids within bacterial membranes, saponification may interfere with pigment identification and yield inconsistent results [1]. Scott et al. [24] demonstrated that sample preparation contributes significantly to overall analytical variability, with saponification conditions—including time, temperature, potassium hydroxide concentration, number of extraction cycles, and initial sample volume, playing critical roles in result accuracy. In our study, reliance on a single saponification protocol may therefore have predisposed the results to methodological bias and partially accounted for the variability observed among isolates.

In the present study, different extraction protocols were not compared, and thus the influence of the selected method on β-carotene quantification remains undetermined. However, the adopted extraction procedure, based on the method described by Radu et al. [25], enabled consistent detection in most analyzed samples. Notably, the absence of detectable β-carotene in MABS confirmed its suitability as a negative reference in method validation, both in visual and chromatographic assessments.

Validation parameters applied to the HPLC-DAD method demonstrated that it is a linear, accurate, and precise approach, with a metabolite recovery rate close to 100%. Moreover, the UHPLC-MS analysis confirmed the specificity of the β-carotene detected, validating its presence even in non-chromogenic species such as MB and MTB. These findings expand our understanding of pigment expression in the Mycobacterium genus and demonstrate the value of chromatographic detection in surpassing the limitations of phenotypic assessments.

Further studies are needed to compare isolates from diverse clinical conditions and determine whether β-carotene production correlates with host–pathogen interactions or disease progression. Additionally, the potential regulatory and protective functions of carotenoids in response to oxidative damage should be investigated, as these compounds may contribute to bacterial persistence in hostile environments. In this study, photochromogenic species were intentionally not exposed to light, as our objective was to quantify basal pigment production under standardized conditions, allowing direct comparison across all species. Nevertheless, we acknowledge that light exposure typically induces higher carotenoid production in these species. Future studies will therefore compare β-carotene levels in photochromogenic isolates under both light-exposed and non-exposed conditions. This comparative approach will not only enhance the robustness of our findings but also position pigment quantification as a potential research tool for exploring light-dependent regulatory mechanisms in mycobacteria, rather than as an immediate diagnostic strategy.

Building upon this premise, the photoinducible behavior observed in certain species has been linked to the transcriptional activation of the crt operon upon light exposure, particularly in MM and MK. In their comprehensive review, Robledo et al. [3] compiled molecular and genetic evidence suggesting that these photochromogenic species possess a conserved crtE–crtB–crtI–crtY gene cluster, whose expression is modulated by light through a phytochrome-like regulatory system. Although their analysis did not include experimental validation, it established a conceptual framework in which β-carotene synthesis in photochromogenic mycobacteria is understood as a genetically mediated, light-responsive process rather than as passive pigment accumulation.

In line with this model, complementary biochemical studies on MK revealed that this species predominantly synthesizes α-carotene (~85%) and lycopene (~15%) as its main carotenoids, while other derivatives appear only in trace amounts [11,26]. This compositional pattern indicates a highly efficient conversion of precursor molecules into terminal products, namely α- and β-carotene, and possibly leprotene. Among these, β-carotene emerges as the principal end product and the pigment most directly involved in photooxidative protection. These findings collectively support the notion that light-induced carotenoid biosynthesis in MK constitutes a genetically regulated adaptive response, integrating environmental cues with metabolic efficiency to strengthen bacterial resilience under oxidative stress conditions.

Expanding on these biochemical insights, recent genomic and transcriptomic analyses have refined the understanding of carotenoid biosynthesis regulation in photochromogenic mycobacteria. According to Janisch et al. [27], two distinct crt gene clusters (crtE–crtB–crtI–crtY and crtW–crtZ) are differentially expressed under light exposure, indicating the existence of specialized biosynthetic routes for various carotenoid derivatives. The study also identified a putative photoreceptor-regulated network that controls crt transcription through a light-inducible, MarR-type transcriptional regulator homologous to CrtR found in other Actinobacteria, suggesting that the light-dependent activation of carotenogenesis in MK involves a dual regulatory system integrating both photoreception and metabolic control.

These findings are consistent with evidence from Henke et al. [28], which demonstrated that CrtR functions as a negative regulator of carotenoid biosynthesis under dark conditions and is inactivated by light-induced conformational changes. The conservation of CrtR-like proteins across Actinobacteria, including mycobacteria, supports the hypothesis that photochromogenic species such as MK utilize analogous regulatory pathways for environmental adaptation, linking gene expression to photic stimuli.

Furthermore, comparative genomic insights from Giraud et al. [29] provide additional support for this model. The study showed that the presence of multiple crt operons enables functional diversification of carotenoid biosynthesis, allowing bacteria to synthesize pigments with distinct photoprotective and structural roles. This evolutionary redundancy may also occur in mycobacteria, where maintaining parallel crt clusters could provide metabolic flexibility and enhanced adaptability to fluctuating light or oxidative conditions.

Beyond its function as a visual marker, β-carotene may play a protective role in bacterial physiology, particularly under oxidative stress conditions. In other pathogenic bacteria, such as Staphylococcus aureus, carotenoids have been shown to neutralize reactive oxygen species (ROS) generated by host immune cells, enhancing bacterial survival within macrophages [30]. Although this function has not been fully characterized in mycobacteria, the detection of β-carotene in species such as MTB and MB, both capable of intracellular persistence, raises the possibility that carotenoid biosynthesis contributes to redox homeostasis during infection. This is consistent with previous findings indicating that pigment-deficient mutants are more susceptible to oxidative damage [31], suggesting a potential adaptive role for carotenoids in the mycobacterial response to host defenses.

The detection of β-carotene in both chromogenic and non-chromogenic mycobacteria using HPLC-DAD and UHPLC-MS suggests that carotenoid profiling represents a valuable research and taxonomic tool, particularly for differentiating closely related NTM species. In species such as MGOR, MK, and MM, pigmentation is traditionally used for classification under the Runyon system. However, as demonstrated in this study, pigment production can vary between isolates and may not always be visually detectable under standard laboratory conditions. This intra-species variability, together with basal carotenoid expression in non-chromogenic species like MTB and MB, underscores the limitations of visual classification systems and highlights the use of chromatographic carotenoid quantification as a complementary approach for species characterization and metabolic assessment.

In conclusion, this study introduces a robust, sensitive, and reproducible analytical approach that reveals low-level β-carotene production in mycobacteria, extending beyond traditional chromogenic classifications. By demonstrating both inter- and intra-species variability and confirming the presence of carotenoids in species historically considered non-pigmented, these findings refine the current phenotypic understanding of Mycobacterium spp. Rather than proposing an immediate diagnostic application, this method should be viewed as a research tool to investigate the biochemical diversity, metabolic adaptation, and potential photoprotective functions of carotenoids across mycobacterial species. The observed heterogeneity in pigment expression further suggests that carotenoid biosynthesis may reflect distinct regulatory mechanisms and adaptive responses to environmental cues, offering new avenues for exploring metabolic regulation and evolutionary differentiation within the genus Mycobacterium.

This study has several limitations that should be considered when interpreting the findings. First, the isolates analyzed were obtained exclusively from the strain collection of the Bacteriology and Mycobacteria Unit at the Corporación para Investigaciones Biológicas (CIB) in Medellín, Colombia. Consequently, the results may not capture the full intra-species variability that exists among geographically distinct populations. Future investigations should include strains from different regions and ecological contexts to validate these observations.

Second, all cultures were grown under dark conditions to ensure standardized comparison across species. Although photochromogenic organisms such as MM and MK exhibit basal carotenoid production even in the absence of light, this design precluded evaluation of light-induced pigment biosynthesis. Further studies incorporating light–dark experimental contrasts are warranted to elucidate the dynamics of photoinducible regulatory systems under physiologically relevant conditions.

Third, carotenoid extraction was performed using a single saponification protocol. While saponification is a widely accepted approach in carotenoid analysis, it may alter pigment stability or recovery efficiency, potentially introducing methodological bias. Employing complementary extraction methods or non-saponifying techniques in future studies could help confirm the reproducibility and accuracy of pigment quantification.

Finally, although the analytical method proved sensitive for detecting β-carotene at low concentrations and revealed carotenoid presence in species traditionally regarded as non-pigmented, its practical implementation currently requires relatively high bacterial biomass. Culturing an isolate remains necessary for reliable pigment detection in clinical specimens.