Methylene blue dose-dependently induces cutaneous inflammation and heat hyperalgesia in a novel rat model

Abstract

Methylene blue (MB) has been shown to reduce mortality and morbidity in vasoplegic patients after cardiac surgery. Though MB is considered to be safe, extravasation of MB leading to cutaneous toxicity has been reported. In this study, we sought to characterize MB-induced cutaneous toxicity and investigate the underlying mechanisms. To induce MB-induced cutaneous toxicity, we injected 64 adult male Sprague-Dawley rates with 200 µL saline (vehicle) or 1%, 0.1%, or 0.01% MB in the plantar hind paws. Paw swelling, skin histologic changes, and heat and mechanical hyperalgesia were measured. Injection of 1%, but not 0.1% or 0.01% MB, produced significant paw swelling compared to saline. Injection of 1% MB produced heat hyperalgesia but not mechanical hyperalgesia. Pain behaviors were unchanged following injections of 0.1% or 0.01% MB. Global transcriptomic analysis by RNAseq identified 117 differentially expressed genes (111 upregulated, 6 downregulated). Ingenuity Pathway Analysis showed an increased quantity of leukocytes, increased lipids, and decreased apoptosis of myeloid cells and phagocytes with activation of IL-1β and Fos as the two major regulatory hubs. qPCR showed a 16-fold increase in IL-6 mRNA. Thus, using a novel rat model of MB-induced cutaneous toxicity, we show that infiltration of 1% MB into cutaneous tissue causes a dose-dependent pro-inflammatory response, highlighting potential roles of IL-6, IL-1β, and Fos. Thus, anesthesiologists should administer dilute MB intravenously through peripheral venous catheters. Higher concentrations of MB (1%) should be administered through a central venous catheter to minimize the risk of cutaneous toxicity.

Introduction

Vasoplegia occurs after 9%–44% of cardiac surgeries. ¹ Conventional treatment of vasoplegia includes administration of fluids and vasopressors, such as phenylephrine, norepinephrine, and vasopressin. Vasopressor therapy can have serious adverse effects, including ischemia of the extremities and intestinal hypoperfusion, which could lead to tissue necrosis and metabolic acidosis. ² Therefore, alternative treatments for vasoplegia have been investigated.

Methylene blue (MB) is a cationic thiazine drug with a characteristic deep blue color that has been used in various clinical applications, including in patients with vasoplegia following cardiac surgery. Early administration of MB was associated with reduced mortality and adverse events, including renal failure. ³ Although MB is generally considered to be safe, several case reports and case series have shown cutaneous inflammation and tissue necrosis MB after accidental extravasation of MB, when given through peripheral intravenous catheters.^4,5 Moreover, MB-induced cutaneous toxicity has been reported in the literature after subcutaneous injections. A retrospective study of 24 patients who were given MB (intradermal 3–5 mL of 1% solution) for sentinel lymph node biopsy reported that 5 patients developed erythematous lesions, necrotic lesions, or superficial ulceration. ⁶ Of 399 cases of sentinel lymph node biopsies performed with MB, 21 cases were complicated with local inflammation characterized by focal erythema and/or palpable induration, and 5 cases resulted in skin necrosis. ⁷ In another study, of 34 patients undergoing sentinel lymph node biopsy with 1% MB for immediate breast reconstruction following mastectomy, 6 patients suffered from complications that included skin necrosis, wound dehiscence, and infection. ⁸ Furthermore, in a retrospective study of 95 patients undergoing lymphatic mapping/sentinel lymphadenectomy for melanoma and breast cancer, 6 patients developed erythema with 2 patients having injection site telangiectasia and 2 patients having fat necrosis. ⁹

Interestingly, MB has also been used to reduce pain. The intracutaneous injection of MB have been used for postherpetic neuralgia, ¹⁰ post-operative pain,^11–13, pruritus ani,^14,15 and various vulvar non-neoplastic epithelial disorders. ¹⁶ Perianal intradermal injection of MB has been used to reduce initial postoperative pain in open hemorrhoidectomies. ¹² and other anorectal surgeries. ¹³ Previous animal studies on cutaneous administration of MB have focused on its therapeutic potential in dermatologic conditions such as burns and wound healing.^17–19 Moreover, Lee et al. found that low doses of MB (40 µL of 0.02%, 0.01%, 0.005% administered subcutaneously) attenuated activity of primary afferent nerve fibers in a dose dependent manner and decreased mechanical and heat hyperalgesia. ²⁰ Thus, we hypothesize that cutaneous administration of MB has opposing effects at low and high concentrations.

Given the widespread use of this drug (266 clinical trials registered on clinicaltrials.gov to investigate the clinical utility of MB in various conditions), it is important to elucidate mechanisms and doses that cause cutaneous toxicity of MB. The aims of this basic science study are to 1) characterize MB-induced cutaneous toxicity in an animal model, 2) identify the dose of MB that cause cutaneous toxicity, and 3) obtain greater insight into the mechanisms underlying MB-induced toxicity.

Materials and methods

Ethics

This study adhered to the guidelines of the Committee for Research and Ethical Issues of the International Association for the Study of Pain and all procedures were approved by the University of Minnesota Institutional Animal Care and Use Committee.

Animals

Adult male Sprague-Dawley rats (250–300 g) were obtained from Harlan (Somerville, NJ). Rats were housed in pairs in polymethyl methacrylate cages measuring 43 × 21.5 × 25.5 cm and maintained on a 12-h light/dark cycle. Food and water were available ad libitum.

Plantar MB injection

Prior to plantar injection, rats were anesthetized with isoflurane. Each rat was placed in a plexiglass induction chamber containing 5% isoflurane in room air. After the righting reflex was lost, 2%-3% isoflurane in room air was delivered through a nose cone. To model MB-induced cutaneous toxicity, rats received an intraplantar injection of saline (vehicle), 1%, 0.1%, or 0.01% MB in a volume of 200 µL into the right hind paw. The highest MB concentration (1%) used in this study is the most used in several clinical settings and is therefore clinically relevant. ²¹ Injections were made using a sterile 1 mL syringe with a 27-G, ½-inch needle. Given the visible nature of MB injection, the investigators could not be blinded to the treatment condition and dosage.

Plethysmography

Rat paw edema following saline, or MB injection was measured as previously described. ²² Briefly, a glass beaker filled with water was tared on a balance, and the rat’s injected hind paw was dipped in the water up to the heel. The displaced water creates a force on the scale, which was recorded. Each paw was measured in triplicate, and data is expressed as an average of three measurements. Measurements were made 4 h, 1 day, and 2 days after injection of saline or 1%, 0.1%, or 0.01% MB (n = 6 per group).

Plantar hind paw skin biopsy

Skin biopsies were collected from the injection site area of the hind paws of rats (n = 5 per group) 24 h after injection of saline, or 1%, 0.1%, or 0.01% MB. Rats were deeply anesthetized with intraparietal injection of Euthasol (390 mg/mL). The hind paws were removed and maintained in Zamboni’s fixative (Newcomer’s Supply, WI, USA) for 24 h 4°C and transferred to 30% sucrose for an additional 24 h at 4°C. A biopsy tool was used to remove a 3 mm biopsy punch from the plantar skin injection site. For sectioning, tissues were immersed in Cryo-gel OCT compound (Electron Microscopy Sciences, Hatfield, PA, USA) and cut into 50-µm thick sections.

Hematoxylin and eosin staining

To investigate histologic changes, hematoxylin and eosin staining was conducted according to routine protocols from hind paw skin biopsies after injection with saline or 1% MB (n = 3 per group). Briefly, following deparaffinization and rehydration, 5-μm longitudinal sections were stained in hematoxylin solution for 5 min, dipped in 1% acid ethanol (1% HCl in 70% ethanol) 5 times, and rinsed in distilled water. Sections were then stained in eosin solution for 3 min, dehydrated with graded alcohol, and cleared with xylene. Mounted slides were examined and photographed using an Olympus BX53 fluorescence microscope (Tokyo, Japan).

Measurement of mechanical hyperalgesia using the electronic von frey test

Rats were placed in individual Plexiglass containers measuring 12 × 20 × 17 cm on top of a metal mesh floor and allowed to acclimate for 30 min. Paw withdrawal thresholds (PWTs), or the threshold force in grams required for paw withdrawal, were measured using an electronic von Frey aesthesiometer (IITC Life Science, Woodland Hills, CA, USA). After rats had acclimated to the testing environment, PWT was determined beginning with a 2 g attachable 0.8 mm polypropylene tip. The tip was pressed perpendicularly against the plantar right hind paw where the injection was administered. Force was applied for 2 s. If there was no response, the next rat would be tested. If there was a response, one more trial was administered to validate the first response. The protocol was repeated using filaments of increasing rigidity. The reported PWT is the average of triplicate measurements. PWT was determined for 3 consecutive days before injection, at 2 h, and at 1, 2, 5, 7, and 9 days after injection of saline, or 1%, 0.1%, or 0.01% MB (n = 6 per group).

Measurement of heat hyperalgesia using the hargreaves test

Following PWT measurements, the same rats were tested for sensitivity to heat by determining the paw withdrawal latency (PWL) to noxious heat. Rats were placed under Plexiglass boxes measuring 12 × 20 × 17 cm on an elevated tempered glass floor of 3 mm thickness. Rats were allowed to acclimate for 20 min, and the floor was maintained at 29.0 ± 1°C using convective heat. A focused radiant heat source (50 W projector lamp with an aperture diameter of 6 mm) underneath the glass floor was aimed at the plantar surface of the injured paw and withdrawal latencies were measured to the nearest hundredths of a second by an automated system. The intensity of the heat was adjusted to produce withdrawal latencies of 10–12 s in naïve rats. PWL was defined as the average of three trials. PWL measurements were conducted 3, 2, and 1 day before injection, and 2 h, and 1, 2, 5, 7, and 9 days after injection.

RNA sequencing

RNA sequencing (RNAseq) was employed to identify global changes in gene expression that covers the entire transcriptome present in the skin. This approach allows for a better resolution of the complex transcriptome including a detailed and quantitative measure of gene expression. Hind paw skin tissues were collected from rats who were injected with saline (n = 2) or MB 1% (n = 2). RNA extraction was performed with RNeasy Plus Universal Mini Kit (QIAGEN Inc.) using the manufacturer’s recommended protocol. Total RNA was isolated, qualified (RIN score >8.0), and subjected to TruSeq stranded mRNA library preparation for rat total RNA Gold library preparation (with ribosomal RNA depletion), according to the manufacturer protocol (Illumina). Quality control was performed for RNA extraction and cDNA library preparation steps with Qubit (Invitrogen) and High Sensitivity NGS fragment analysis kit on the Fragment Analyzer (Agilent Technologies). After standardizing the amount of cDNA per sample, the libraries were sequenced on an Illumina NextSeq500 sequencing platform with 75-bp single-end reads in multiplexed sequencing experiments, yielding a median of 22.3 million reads per sample. mRNA library preparation and sequencing were done at the University of Minnesota Genomic Core Facility.

Bioinformatics

NGS reads were trimmed with Trimmomatic (v 0.33) enabled with the “-q” option; 3 bp sliding-window trimming from 3′ end requiring minimum Q30. Quality control on raw sequence data for each sample was performed with FastQC. Read mapping was performed with Hisat2 (v2.1.0) ²³ with the rat genome (rn6) as a reference. Differentially expressed genes (DEGs) were identified using edgeR (v. 3.42.4). Raw counts were normalized using the Trimmed Mean of M-values (TMM) method. TMM-normalized transcript counts were fitted to a general linear model to identify DEGs. DEGs (≥1.5x Absolute Fold Change and a false discovery rate (FDR) corrected p < .05) were analyzed by Ingenuity Pathway Analysis (IPA, Qiagen, USA), which consists of 8.5 million findings, to identify MB-altered gene networks and associated biological functions using the “Grow” feature as previously described. ²⁴

Quantitative real-time polymerase chain reaction (qPCR)

qPCR was performed to analyze specific pro-inflammatory mRNA expression changes in the hind paw skin 2-h after treatment. Hind paw skin tissues were collected from rats injected with saline (n = 5) or MB 1% (n = 5). One mg of total RNA isolated from the skin (RNAqueous RNA isolation kit, Life Technologies Inc.) was used to generate complementary DNA (High-Capacity cDNA reverse transcription kit, Life Technologies Inc.). Relative expression levels were quantified for inflammatory markers interleukin-6 (IL-6), allograft inflammatory factor (AIF)-1, and tumor necrosis factor (TNF)-α using TaqMan Gene Expression Assays (Applied Biosystems, Foster City, CA) and a DNA analyzer (Quant Studio, Applied Biosystems Inc.). Relative fold changes are means of biological replicates, each performed in technical quadruplicate, normalized to beta-actin (internal loading control), and transformed to be relative to saline-injected hind paws.

Statistical analysis

Two-way analyses of variance with repeated measures were used to compare the differences in PWL, PWT, and paw edema measurements between saline and MB 1% injected rats. Unpaired t-tests were used to compare mRNA expression level changes between saline and 1% MB injected rats. Analyses were performed using Prism 9 (GraphPad Software, San Diego, CA). A p value of <0.05 was considered significant. Data are presented as means ± SEM.

Results

Dose-dependent paw swelling and epidermal tissue changes following MB injection

Thirty minutes following injection of 200 µL of 1% MB into the rat plantar hind paw skin, paw edema was evident (Figure 1(a)). To quantify the paw edema over time, plethysmography was performed 2 h, 1 day, and 2 days following injection. Paw swelling was increased following 1% MB injection when compared to saline over the measurement period (Two-way ANOVA F (2, 20) = 4.501, p = .0243). The greatest hind paw swelling following 1% MB infiltration occurred 2 h after injection (2.9 ± 0.4 mL) which was significantly greater than control paws 2 h after saline injection (2.1 ± 0.2 mL, p < .0001; Figure 1(b)). Compared to saline, hind paw swelling persisted for 1 day (1% MB 2.7 ± 0.3 mL, saline 2.0 ± 0.1 mL, p < .0001) and 2 days after injection (1% MB 2.4 ± 0.2 mL, saline 2.0 ± 0.1 mL, p = .0149; Figure 1(b)). There were no differences in paw swelling following saline, 0.1% MB or 0.01% MB at any time (Figure 1(c)).

Figure 1.

Paw swelling following MB skin infiltration. (a) Representative image of macroscopic changes in plantar surface of rat right hind paw 30 min after infiltration of 200 µL of saline (left) or 1% MB (right). (b)-(c) Paw swelling measured at 2 h, 1 day, and 2 days after plantar hind paw skin injection. (b) Saline (n = 6) and 1% MB (n = 6) injected rats are graphed. (c) Saline (n = 6), 0.1% MB (n = 6), and 0.01% MB (n = 6) injected rats are graphed. The horizontal axis is not scaled linearly with time. Markers represent mean values. Error bars represent SEM. Asterisks denote significant differences (****p < .0001, *p < .05).

To examine the histological changes in the epidermis following MB injection, hematoxylin-eosin staining was performed. Skin biopsies taken 1 day after injection of 1% MB injection showed increased epidermal thickness and neutrophilic infiltration (Figure 2).

Figure 2.

Representative hematoxylin and eosin-stained skin sections near saline or 1% MB injection sites. Epidermis thickness is indicated by orange double headed arrows. Increased infiltration of neutrophils in MB-injected skin sections are circumscribed with an orange line.

1% MB produces heat but not mechanical hyperalgesia

Heat hyperalgesia was measured at before and at 2 h, and 1, 2, 5, 7, and 9 days after intraplantar injection of saline, 1%, 0.1% or 0.01% MB (n = 6 per group). Injection of 1% MB decreased PWL compared to saline or the lower concentrations of MB (Two-way ANOVA F (6, 96) = 5.852, p < .0001). PWL was significantly decreased at 2 h (1% MB 7.250 ± 0.486, saline 11.909 ± 0.608 s, p < .0001) and 1 day (1% MB 10.128 ± 0.457, saline 12.360 ± 0.719 s, p = .0221) after 1% MB injection compared to saline, before returning to baseline (Figure 3(a)). PWL was not altered following 0.1% and 0.01% MB and did not differ from the saline-treated group (Figure 3(b)).

Figure 3.

Pain behaviors following subcutaneous injection of 200 µL of saline or 1%, 0.1%, or 0.01% MB. (a)-(b) Heat hyperalgesia was measured using PWL at baseline, 2 h, 1 day, 2 days, 5 days, 7 days, and 9 days after injection. (a) Saline (n = 6) and 1% MB (n = 6) injected rats are graphed. (b) Saline (n = 6), 0.1% MB (n = 6), and 0.01% MB (n = 6) injected rats are graphed. (c)-(d) Evoked mechanical allodynia was using PWT at baseline, 2 h, 1 day, 2 days, 5 days, and 7 days after injection. (c) Saline (n = 6) and 1% MB (n = 6) injected rats are graphed. (d) Saline (n = 6), 0.1% MB (n = 6), and 0.01% MB (n = 6) injected rats are graphed. The horizontal axis is not scaled linearly with time. Markers represent mean values. Error bars represent SEM. Asterisks denote significant differences (****p < .0001, *p < .05).

Unlike hyperalgesia to heat, mechanical hyperalgesia did not develop following MB. No differences in PWT were identified following injection of saline or MB at any time (Figure 3(c) and (d)).

Transcriptomic changes following injection of 1% MB

To identify genes that potentially contribute to MB-induced paw swelling, epidermal structural changes, and heat hyperalgesia, we investigated global transcriptomic changes in the hind paw plantar skin following MB injection. Using the selection criteria (a minimum 1.5x Absolute Fold Change and an FDR <0.05), 117 DEGs (111 up- and 6 down-regulated) were identified between MB- versus saline-treated rats (Table 1). IPA mapped these genes onto specific diseases and biological functions centering on activation of the interleukin-1 beta (IL-1β) and Fos gene networks (Figure 4). Compared to saline, MB changed expression of gene networks associated with an increased quantity of leukocytes, increased lipids, and decreased apoptosis of myeloid cells and phagocytes.

Table 1.

List of significantly differentially expressed genes in plantar hind paw skin after 1% methylene blue infiltration.

Gene name	Gene symbol	log2 (fold change)	FDR
Ubiquitin specific peptidase 27, X-linked	Usp27x	10.5	5.90 × 10−11
Apoptosis antagonizing transcription factor	Aatf	10.4	5.90 × 10−11
Similar to heat shock protein 8	LOC680121	10.4	1.41 × 10−08
Phosphatidylinositol glycan anchor biosynthesis, class W	Pigw	9.93	4.90 × 10−06
Rhox homeobox family member 11	Rhox11	9.66	4.71 × 10−07
Uncharacterized LOC102552565	LOC102552565	9.57	2.60 × 10−06
NAD synthetase 1	Nadsyn1	9.30	1.07 × 10−04
Zinc finger protein 454	Znf454	8.45	1.47 × 10−03
Somatostatin receptor 2	Sstr2	8.12	4.16 × 10−17
Small nucleolar RNA SNORA72	LOC120095384	8.10	9.98 × 10−14
microRNA 27a	Mir27a	8.07	8.18 × 10−03
Retinol dehydrogenase 11	Rdh11	7.66	3.48 × 10−02
Phosphodiesterase 8B	Pde8b	7.50	1.58 × 10−25
ADP-dependent glucokinase	Adpgk	7.16	8.02 × 10−04
Ankyrin repeat and zinc finger peptidyl tRNA hydrolase 1	Ankzf1	6.38	4.28 × 10−07
Spermatogenesis associated 3	Spata3	6.29	2.23 × 10−04
Similar to Ly6-C antigen gene	RGD1565410	6.06	1.38 × 10−03
Septin 9	Septin9	6.05	2.61 × 10−21
G protein nucleolar 2	Gnl2	5.63	4.09 × 10−02
Variable coding sequence A2	Vcsa2	5.63	1.58 × 10−02
Switching B-cell complex subunit SWAP70	Swap70	5.56	2.19 × 10−03
Brain-enriched guanylate kinase-associated	Begain	5.53	2.70 × 10−07
Tubulin polymerization promoting protein	Tppp	5.51	5.53 × 10−10
Prostate and testis expressed protein 2-like	LOC103693040	5.34	7.69 × 10−03
PH domain and leucine rich repeat protein phosphatase 1	Phlpp1	4.82	1.90 × 10−06
Growth hormone 1	Gh1	4.79	2.34 × 10−08
Serpin family C member 1	Serpinc1	4.66	4.22 × 10−02
Alcohol dehydrogenase 6A (class V)	Adh6a	4.61	8.45 × 10−10
Neurofilament medium chain	Nefm	4.61	1.58 × 10−02
Cytoskeleton-associated protein 4	Ckap4	4.58	1.22 × 10−08
Protein geranylgeranyltransferase type I subunit beta	Pggt1b	4.47	9.28 × 10−17
Lysine acetyltransferase 14	Kat14	4.43	2.64 × 10−18
Defensin beta 33	Defb33	4.36	3.62 × 10−02
Similar to ribosomal protein L10a	LOC680700	4.34	2.41 × 10−08
Adaptor related protein complex 1 subunit sigma 1	Ap1s1	4.28	2.73 × 10−10
Interferon-induced protein with tetratricopeptide repeats 2	Ifit2	4.20	6.70 × 10−04
CD3 delta subunit of T-cell receptor complex	Cd3d	4.17	6.99 × 10−03
Solute carrier family 50 member 1	Slc50a1	4.14	2.30 × 10−09
Phospholipase A2, group IVF	Pla2g4f	3.99	1.58 × 10−06
Salt-inducible kinase 1	Sik1	3.95	3.23 × 10−08
U1 spliceosomal RNA	LOC120102701	3.94	2.46 × 10−05
Focadhesin	Focad	3.92	2.19 × 10−03
U2 spliceosomal RNA	LOC120098022	3.86	5.17 × 10−04
Dehydrogenase/reductase 1	Dhrs1	3.73	1.46 × 10−03
Golgi associated kinase 1B	Gask1b	3.64	9.38 × 10−10
PDZ and LIM domain 4	Pdlim4	3.61	4.71 × 10−07
PDZ domain containing RING finger 4	Pdzrn4	3.60	3.23 × 10−08
Teratocarcinoma-derived growth factor 1	Tdgf1	3.59	1.54 × 10−03
JAZF zinc finger 1	Jazf1	3.56	7.34 × 10−06
Dermatopontin	Dpt	3.54	1.71 × 10−05
Dishevelled segment polarity protein 1	Dvl1	3.54	1.53 × 10−10
Sialophorin	Spn	3.53	2.76 × 10−04
mutS homolog 5	Msh5	3.53	1.34 × 10−02
Pyruvate kinase M1/2, pseudogene 20	Pkm-ps20	3.47	4.37 × 10−02
Asparagine synthetase (glutamine-hydrolyzing)	Asns	3.36	3.71 × 10−06
MOB kinase activator 1B	Mob1b	3.20	9.23 × 10−07
Notch receptor 4	Notch4	3.18	6.54 × 10−03
Cache domain containing 1	Cachd1	3.18	4.38 × 10−03
Rhomboid domain containing 3	Rhbdd3	3.17	4.92 × 10−02
Tudor domain containing 6	Tdrd6	3.14	4.16 × 10−03
Epsin 1	Epn1	3.14	1.52 × 10−02
Retinol dehydrogenase 8	Rdh8	3.13	2.61 × 10−07
Ubiquinol-cytochrome c reductase core protein 1	Uqcrc1	3.11	1.36 × 10−02
Coilin	Coil	3.11	5.90 × 10−11
Testis expressed 46	Tex46	3.09	2.98 × 10−04
Leiomodin 2	Lmod2	3.08	6.21 × 10−05
RAP2C, member of RAS oncogene family	Rap2c	3.06	5.85 × 10−03
Small integral membrane protein 12	Smim12	3.05	7.10 × 10−05
SWI5 homologous recombination repair protein	Swi5	3.02	1.34 × 10−02
NADH:ubiquinone oxidoreductase subunit S5	Ndufs5	2.99	7.23 × 10−07
Nitric oxide synthase 3	Nos3	2.92	2.07 × 10−04
Envoplakin	Evpl	2.90	3.57 × 10−03
LON peptidase N-terminal domain and ring finger 3	Lonrf3	2.90	3.06 × 10−06
Transmembrane protein 109	Tmem109	2.90	3.96 × 10−04
BAF chromatin remodeling complex subunit BCL7C	Bcl7c	2.90	6.56 × 10−04
Family with sequence similarity 107, member A	Fam107a	2.77	7.26 × 10−03
Acyl-CoA dehydrogenase, long chain	Acadl	2.76	1.03 × 10−03
ASXL transcriptional regulator 3	Asxl3	2.71	7.42 × 10−07
Kelch domain containing 9	Klhdc9	2.70	7.26 × 10−03
Leucine rich repeat containing 23	Lrrc23	2.68	5.34 × 10−04
Myosin IIIB	Myo3b	2.68	9.32 × 10−03
Adhesion G protein-coupled receptor A1	Adgra1	2.67	1.76 × 10−03
Mitochondrial translation release factor 1 like	Mtrf1l	2.62	1.56 × 10−04
Adrenoceptor alpha 2B	Adra2b	2.55	4.56 × 10−02
Spectrin repeat containing, nuclear envelope family member 3	Syne3	2.52	6.21 × 10−05
Caspase 12	Casp12	2.51	8.05 × 10−06
DExD-box helicase 21	Ddx21	2.48	1.55 × 10−05
RFT1 homolog	Rft1	2.46	2.21 × 10−02
Calponin 3	Cnn3	2.41	2.43 × 10−06
NHS-like 2	Nhsl2	2.41	5.17 × 10−04
trans-golgi network vesicle protein 23 homolog B	Tvp23b	2.40	2.07 × 10−04
N (alpha)-acetyltransferase 38, NatC auxiliary subunit	Naa38	2.32	2.66 × 10−04
Neurexophilin and PC-esterase domain family, member 3	Nxpe3	2.28	1.55 × 10−03
Tnf receptor-associated factor 3	Traf3	2.26	2.43 × 10−06
Ufm1-specific ligase 1	Ufl1	2.22	3.52 × 10−04
Component of oligomeric golgi complex 1	Cog1	2.18	2.68 × 10−03
Kelch-like family member 4	Klhl4	2.12	5.85 × 10−05
Cytidine and dCMP deaminase domain containing 1	Cdadc1	2.11	3.32 × 10−02
Mitotic arrest deficient 2 like 2	Mad2l2	2.10	2.56 × 10−05
REC114 meiotic recombination protein	Rec114	2.10	2.17 × 10−02
Argininosuccinate lyase	Asl	2.03	1.85 × 10−04
Zinc finger protein 105	Zfp105	2.00	6.39 × 10−05
Tripartite motif containing 63	Trim63	1.97	2.74 × 10−03
NSF attachment protein gamma	Napg	1.96	2.18 × 10−04
H1.7 linker histone	H1f7	1.94	3.29 × 10−03
Cellular repressor of E1A-stimulated genes 2	Creg2	1.93	1.30 × 10−03
SLX9 ribosome biogenesis factor	Slx9	1.93	2.01 × 10−02
Jumping translocation breakpoint	Jtb	1.83	2.84 × 10−02
Peptidylprolyl isomerase like 3	Ppil3	1.69	1.25 × 10−03
Folliculin	Flcn	1.64	1.93 × 10−02
Ly6/Plaur domain containing 5	Lypd5	1.64	1.65 × 10−02
Ribosomal RNA processing 15 homolog	Rrp15	−1.54	3.25 × 10−02
Collagen type VIII alpha 1 chain	Col8a1	−1.74	2.60 × 10−02
Keratin 15	Krt15	−1.76	6.94 × 10−03
Major facilitator superfamily domain containing 6	Mfsd6	−1.80	1.38 × 10−03
Fc fragment of IgG binding protein-like 1	Fcgbpl1	−2.81	1.99 × 10−03
Yippee-like 3	Ypel3	−8.00	2.05 × 10−02

Figure 4.

Altered gene regulatory networks and associated downstream effects in rat hind paw plantar skin following methylene blue 1% injection. Changes in gene expression lead to increased quantity of leukocytes, increased lipids, and decreased apoptosis of myeloid cells and phagocytes. Predicted activation is in orange, and predicted inhibition is in blue.

Given the evidence of inflammation due to increased leukocytosis and persistence of inflammatory cells, we next used qPCR to investigate the expression of several inflammatory mediators. IL-6, AIF-1, and TNF-α mRNA levels were compared in skin biopsies following 1% MB and saline treatment. IL-6 expression was elevated in 1% MB skin biopsies compared to saline (16.8 ± 4.0 and 1.0 ± 0.5, respectively, p = .0039, Figure 5(a)). AIF-1 and TNF-α mRNA levels were not significantly different (Figure 5(b) and (c)).

Figure 5.

Rat hind paw skin mRNA expression changes following subcutaneous injection of 1% MB (n = 5) relative to saline (n = 5). (a) AIF-1, (b) IL6, and (c) TNF- α. Bars represent mean and SEM. Asterisks denote significant differences (**p < .01, ns = not significant).

Discussion

Our study is the first to describe MB-induced cutaneous toxicity in an animal model. Our results show that cutaneous infiltration of a clinically relevant dose of MB (1%) produced a concentration-dependent inflammatory reaction as evidenced by paw edema, increased epidermal thickness, heat hyperalgesia, a 16-fold increase of pro-inflammatory mRNA IL-6, and activation of IL-1β and Fos signaling pathways. Our results suggest that infiltration of 1% MB into cutaneous tissue is contraindicated. At lower concentrations (0.1% or 0.01%), MB did not induce inflammatory responses of paw swelling and pain behaviors, suggesting a dose-dependent toxicity. Although lower concentrations of MB are safer, their clinical efficacy needs to be investigated. One study showed that diluted MB (<1%) results in fewer complications without reducing mapping success. ²⁵ However, for most clinical applications, 1% MB is typically used for tissue infiltration. ¹⁸ Our results agree with several clinical studies that showed cutaneous administration of MB produced skin erythema, ulceration, and necrosis.^6,9

Although MB-induced skin toxicity is well-documented in clinical reports, the precise mechanisms remain unknown.^6–9 MB is known to be oxidized in the body to form formaldehyde and other deaminized oxide radicals. ²⁶ An excessive amount of these products, particularly in tissues that are poorly perfused, may be engulfed by macrophages and cause local inflammation. The tissue reaction to MB is a foreign body reaction characterized by eosinophilic infiltration and fibrinoid necrosis with ischemic ulceration. We assessed global gene expression changes following 1% MB injection, finding 117 DEGs in MB- versus saline-injected rats. IPA suggested an activation of an IL-1β pro-inflammatory mechanism with leukocytosis and increased persistence of myeloid cells and phagocytes. Consistent with these findings, expression of the pro-inflammatory marker IL-6, but not AIF-1 or TNF-α, was elevated in MB-injected paws, supporting an IL-6 mediated inflammatory reaction.

In contrast to our results, analgesic effects have been reported after MB-induced cutaneous infiltration for hemorrhoid and other anorectal surgeries, and pruritus ani.^11–15 It is unknown if mixture with local anesthetics can ameliorate MB induced toxicity. In a study of subcutaneous injection of different concentrations of MB after hemorrhoidectomy, 0.1% and 0.2% MB had similar analgesic effects and adverse event profiles. ²⁷ In agreement with these clinical studies, previous studies reported dose-dependent analgesia in rats following lower doses (40 µL of 0.02%, 0.01%, 0.005% administered subcutaneously) of MB. ²⁰ Thus, lower concentrations and doses of MB have analgesic properties. Although we used a lower concentration of MB (0.001%), this was higher than that used in the above studies (200 µL vs 40 µL).

In conclusion, using a novel animal model we show that infiltration of 1% MB into the skin causes skin toxicity with a dose-dependent inflammatory reaction accompanied by edema and heat hyperalgesia. Activation of cytokine-mediated pathways (i.e., IL-1β, IL-6) could underlie these pathological changes. A limitation of our study is that it was conducted in rats; species differences could play a role in the doses required for MB-induced toxicity.

Considering the results presented in this study, use of 1% MB directly into skin (such as, during hemorrhoidectomy or breast surgery) requires caution. Approximately 10%–30% of patients receiving intravenous medications experience extravasation into tissues. ²⁸ Therefore, anesthesiologists should administer dilute MB (<1%) whenever possible. If a higher concentration of MB (1%) is needed, caution should be taken in patients with higher risk factors for extravasation of peripheral venous catheters, such as being under general anesthesia or heavy sedation, inaccessible catheter sites, and improperly placed catheters. Alternatively, MB 1% should preferably be administered through a central venous catheter to minimize the risk of cutaneous toxicity.

ORCID iD

Twan Sia https://orcid.org/0000-0003-3037-8680