OUP user menu

Determination of Serum BRAK/CXCL14 Levels in Healthy Volunteers

Kazuhito Izukuri DDS, PhD, Shin Ito PhD, Naohito Nozaki PhD, Nobuyuki Yajima PhD, Mariko Iwamiya BSc, Sachiko Kawahara BA, Kenji Suzuki DDS, PhD, Eiro Kubota DDS, PhD, Ryu-Ichiro Hata PhD
DOI: http://dx.doi.org/10.1309/LMQOXCQEF7ZXIIUK 478-482 First published online: 1 August 2010


Background: BRAK/CXCL14 is a chemokine expressed in many normal cells and tissues. Several papers have reported tissue distribution, but no study has examined the serum protein level of this chemokine.

Methods: We evaluated serum levels of BRAK/CXCL14 by enzyme-linked immunosorbent assay, after determining the conditions necessary to increase the sensitivity and reproducibility of the method.

Results: The average level of males was significantly higher than that of females. Two percent of the volunteers showed a value 2 S.D. higher than the average value without any obvious abnormality, but no person was lower than 1 S.D. of the average value. The serum BRAK/CXCL14 levels of individual subjects were constant.

Conclusion: These data indicate that the serum BRAK/CXCL14 level is highly conserved and suggest that the levels are determined by the genetic background and/or lifestyle of the individual.

  • chemokine
  • BRAK
  • CXCL14
  • enzyme-linked immunosorbent assay (ELISA)

Chemokines are a group of small proteins with molecular weights in the range of 8 to 12 K. They are mostly basic and structurally related molecules that are reported to regulate cell trafficking of various types of leukocytes. They function by interacting with a subset of 7-transmembrane, G protein-coupled receptors. Chemokine domains are defined by the presence of 4 conserved cysteine residues linked by 2 disulfide bonds. Two major subfamilies, CXC and CC chemokines, are distinguished according to the position of the first 2 cysteine residues, which are separated by 1 amino acid (CXC subfamily) or are adjacent to each other (CC subfamily).1

Chemokine (C-X-C motif) ligand 14 (CXCL14), also known as breast and kidney-expressed chemokine (BRAK) or BRAK/CXCL14, is a non-ELR (ie, the BRAK/CXCL14 chemokine lacks a Glu-Leu-Arg [or ELR, where E, L, and R are the 1-letter abbreviations for the amino acids, glutamic acid, leucine, and arginine, respectively] tripeptide sequence adjacent to the CXC motif) CXC chemokine expressed ubiquitously and constitutively in epithelial tissues throughout the body.2 Several physiological functions of this chemokine have been proposed such as recruitment and maturation of monocyte-derived macrophages and renewal of Langerhans cells in the skin.3,4 Other functions include stimulation of trafficking of activated natural killer cells to the sites of inflammation or malignancy5 and macrophage infiltration into white adipose tissue in obese mice fed a high-fat diet,6 as well as inhibition of angiogenesis.7 Regulation of tumor formation by the chemokine has also been reported.8,9

For the first step to investigate the mechanism of action of BRAK/CXCL14 in vivo in humans, we modified the enzyme-linked immunosorbent assay (ELISA) to determine serum levels of BRAK/CXCL14 protein and applied this modified ELISA to serum samples from healthy volunteers.

Materials and Methods

Study Population and Study Groups

The study population consisted of 846 subjects (516 males and 330 females; ages 20 to 66 years), who were students or employees of Kanagawa Dental College and participated in this study at the time of a regular health examination. Written informed consent was obtained from those who agreed to participate voluntarily, and ethical clearance was obtained from the institution’s ethics committee.

Collection and Preservation of Blood Samples

Five mL of blood was collected from the antecubital fossa by venipuncture using a VENOJECT II blood collection system (Terumo, Tokyo, Japan) composed of a needle (MN-2138MS), a holder (XX-VP010HD), and a 6 mL syringe (VJ-AS076A001). The blood sample was allowed to clot at room temperature for 1 hour, and then the serum was separated from the blood by centrifuging at 3000 × g and 4°C for 5 minutes. The separated serum was immediately transferred to 3 plastic tubes and stored at −30°C until the time of the assay. Ten samples were randomly picked, and a half of each sample was kept at 4°C, with the remaining half kept at −30°C to examine the effects of freezing and thawing on the serum BRAK/CXCL14 value. Blood plasma was also obtained by use of a Vacutainer (Code No. 367845, Becton Dickinson Japan, Tokyo, Japan). In this case, blood samples were immediately separated by centrifugation at 3000 × g and 4°C for 5 minutes.

Determination of BRAK/CXCL14 in Serum Samples

BRAK/CXCL14 levels in serum obtained from the subjects were measured by using a solid-phase sandwich ELISA (Catalog Number: DY866, DuoSet ELISA Development System, human CXCL14/BRAK, R&D Systems, Minneapolis, MN) according to the manufacturer’s recommendations with some modifications, such as using a fluorogenic peroxidase substrate instead of a colorimetric enzyme substrate and changing the concentrations of both capture and detection antibodies and the duration of plate coating to increase sensitivity and reproducibility.

Briefly, 50 μL (instead of 100 μL) of serum samples and standards (recombinant human BRAK/CXCL14, R&D Systems) were incubated for 2 hours at room temperature in 96-well plates (No Lid, Flat Bottom, Costar 9018, Corning, Corning, NY) precoated with primary, capture antibody in the amount of 1 or 2 μg. After incubation, the wells were washed 3 times with phosphate-buffered saline containing 0.05% Tween 20 and then streptavidin-horseradish peroxidase-conjugated polyclonal antibodies (250 or 500 ng) against BRAK/CXCL14 were added to each well. After incubation for 1 hour, instead of 20 minutes at room temperature, QuantaBlu fluorogenic peroxidase substrate solution (Pierce, Rockford, IL) was added. One hour later, the emitted fluorescence was measured with a Spectrafluor (Tecan Trading AG, Männedorf, Switzerland) operated at an excitation wavelength at 325 nm and an emission wavelength at 420 nm. The concentration of BRAK/CXCL14 in the test samples was estimated by referring to a previously prepared calibrated standard curve, which was obtained by plotting the fluorescence value of the standards against their concentration by use of PLATEmanager 2001 software (Wako Pure Chemical Industries, Osaka, Japan).

Statistical Analysis

Welch’s t test was used to assess statistically significant differences in the values between males and females, between the values of individuals, and an average for the same age group, with P<0.05 being considered statistically significant.


Effects of Antibody Concentration on the Fluorescence Value

We could not detect the chemokine by ELISA with a colorimetric method. Thus, we examined conditions to increase the sensitivity of the system. First we employed fluorogenic (QuantaBlu) instead of the colorimetric peroxide substrate, and obtained a 10-fold increase in sensitivity. Next, by increasing the concentration of capture and detection antibodies 2 fold and the coating time 3 fold, we could reproducibly detect as little as 0.05 ng/mL cytokine in the samples and obtained a linear relationship between the BRAK/CXCL14 concentration and fluorescence intensity up to 10 ng/mL (Figure 1).

Examination of Collection Time, Preservation Conditions of Samples, and Recovery of Supplemented Standard

Before beginning a large-scale experiment, we examined circadian variation, such as sample collection in the morning or in the afternoon and before and after food intake as well as day-to-day variations, by using the serum from 3 volunteers. The values obtained were constant regardless of the time of blood collection, and we could not find any significant difference among these conditions of sample collection. Recovery after freezing samples at −30°C and thawing was 96 + 5%. Thus we stocked all the samples at this temperature. The recovery of 10 ng of recombinant BRAK/CXCL14 added to the serum samples was 95.5 ± 1.3 %. Determination of recombinant BRAK/CXCL14 from another company (Pepro-Tec, Hamburg, Germany) and recombinant mouse BMAC/CXCL14 provided the same standard curve (data not shown).

Figure 1

Effects of concentration of capture and detection antibodies on the intensity of fluorescence. (A) BRAK concentration was determined by use of the QuantaBlu fluorogenic peroxidase substrate and a DuoSet ELISA Development kit (R&D). A standard curve with linearity from 0.5 to 3 ng/mL was obtained. (B) Lower limit of detection of BRAK/CXCL14. The standard curve with linearity from 0.05 to 0.4 ng/mL is shown. (C) Relationship between concentrations of BRAK/CXCL14 (1 to 10 ng) and fluorescence intensity. The meaning of each symbol is given in the figure.

In order to investigate the effects of possible contamination of the serum samples by red blood cells and/or leukocytes, we added red blood cell lysate and/or buffy coat containing peripheral blood mononuclear cells obtained from a plasma preparation to serum samples. We could not find any effects on the value of the serum level of BRAK/CXCL14 (data not shown).

Determination of Serum BRAK/CXCL14 Level

All of the samples were measured at least twice, and the differences between 2 determinations were less than 10%; thus simple averages of the 2 determinations were used for the calculation except when indicated otherwise. The mean BRAK/CXCL14 serum concentration for men (516 subjects, 1.14 ± 0.45 ng/mL) was significantly higher than that for women (330 subjects, 0.94 ± 0.21 ng/mL, P<0.001; Table 1, Figure 2A). In addition, the values for males of all age groups from their 20s (20–29) to their 60s (60–66) were significantly higher than those for the females (Table 1, Figure 2A).

Serum samples from 12 males and 6 females gave a value 2 S.D. higher, and 5 males and 4 females, a value 3 S.D. higher, than the respective average values (Figure 2B and Figure 2C). However, no subject had a value lower than 1 S.D. from the average. The lowest value for males was 0.78 ng/mL and for females was 0.71 ng/mL. Both values were within 1 S.D. of the respective average values (Figure 2B and Figure 2C). In order to confirm the reproducibility of the value over time, we obtained blood samples from 7 representative subjects from the 20s to 60s groups at 6 months and/or 3 months after the first sampling and measured the BRAK/CXCL14 levels by 4 to 6 determinations. The serum BRAK/CXCL14 values of the second sampling taken 3 months after the first sampling (b of Figure 3) and third sampling, taken 6 months after the first 1 (c of Figure 3) of the high-concentration group (>2 ng/mL) were the same as the respective values of the first sampling (a of Figure 3), indicating the respective serum BRAK/CXCL14 values to be reproducible and constant.


The chemokine BRAK/CXCL14 is detected in various parts of the body,2 but no data on normal BRAK/CXCL14 protein levels in the blood have been reported because of the absence of an appropriate method for its determination. Recently, a Western blot method was reported for detection of mouse serum BRAK/CXCL14 levels; but in this case, although an increase in the serum BRAK/CXCL14 level was detected in mice fed a high-fat diet, the normal level in the serum was not detectable by this method.10 Therefore, we chose to use the ELISA method to detect BRAK/CXCL14 and to improve its sensitivity by employing fluorescent peroxidase substrate and changing the concentrations of both capture and detection antibodies. The sensitivity of the modified method was increased up to 20 times; and with these modifications we could lower the detection limit to less than 0.05 ng/mL of BRAK/CXCL14 in the serum (Figure 1). The average serum level of BRAK/CXCL14 of male subjects was significantly higher than that of females in all age groups. A greater number of the male subjects showed significantly higher BRAK/CXCL14 protein levels than the female subjects (12 men and 6 women). In order to avoid the apparent difference in the average values of the males and females by the presence of high-level samples, we eliminated values higher than 2 S.D. from the average (12 men and 6 women) from the respective calculation and calculated the average values from the remaining subjects again. The average serum BRAK/CXCL14 level of males (1.09 ± 0.25 ng/mL, 504 men) was still significantly (P<0.001) higher than that of women (0.92 ± 0.11 ng/mL, 324 women). These data indicate that the difference between the sexes was not dependent on the presence of males with high values. The reason for this sex difference is not clear, but sex-dependent difference in the function of BRAK/CXCL14 has also been reported.6

Of the study population, about 2% of the individuals indicated a value 2 S.D. higher than the average value, but no one had a value lower than 1 S.D. from the average (Figure 2B and Figure 2C). In order to examine the reproducibility of the value, we recollected a blood sample from 7 subjects at 3 months and from 6 subjects at 6 months after the first examination and determined their serum BRAK/CXCL14 levels. The values of the high-expressing group were the same as at the first sampling, and redetermination of the first sample after refreezing and thawing showed the same value, indicating stability against freezing/thawing, reproducibility of the determination method, and consistency of the serum BRAK/CXCL14 over time. The results also indicate that BRAK/CXCL14 is classified as an inflammatory chemokine1 and has homeostatic functions. These data further suggest that the high expression level may be due to the genetic background and/or lifestyle of the respective subjects.

View this table:
Table 1

Serum BRAK/CXCL14 Levels of 516 Male and 330 Female Subjects

AgeSexNumber of SubjectsAverage (ng/mL)Standard DeviationPValue
  • Averages of total and subgroups with ages are shown.

Figure 2

Distribution of serum BRAK/CXCL14 levels in male and female subjects. (A) Differences in serum BRAK/CXCL14 levels between male and female subjects according to age. (B) Distribution of individual serum BRAK/CXCL14 levels in male subjects. (C) Distribution of individual serum BRAK/CXCL14 levels in female subjects.



Serum BRAK/CXCL14 levels of wild-type normal mice were also 1 ng/mL blood.11 Transgenic mice genetically manufactured by introducing a human BRAK/CXCL14 expression vector expressed serum BRAK/CXCL14 at a 10 times higher level than wild-type mice, and these mice were resistant to the growth of tumor cell xenografts.11 Interestingly, 1 person in our study showed a high level, 8.43 + 0.20 ng/mL, of serum BRAK/CXCL14 (Figure 3), which is nearly the level found in such transgenic mice. This individual showed no signs of inflammation or any other abnormality as determined by serum biochemical examination (data not shown).

The determination of serum BRAK/CXCL14 levels by the modified method described here may be very useful for the detection of fluctuations of the value in individuals and for clarification of molecular mechanisms underlying the regulation of the serum BRAK/CXCL14 level. Furthermore, this method should be beneficial for the determination of BRAK/CXCL14 levels in various tissues in vivo and in culture media under in vitro conditions.

Figure 3

Reproducibility and constancy of the serum BRAK/CXCL14 levels. a, the first determination; b, samples re-collected 3 months after the first collection; c, samples re-collected 6 months after the first collection. Reasonably constant values were obtained for the respective subjects, especially for those in the high-concentration group. ***P<0.001


We thank Ms. Etsuko Shimada for preparation of the reference list. This work was supported in part by a grant-in-aid from the High-Tech Research Center Project of the Ministry of Education, Culture, Sports, Science, and Technology of Japan and by a grant-in-aid for scientific research from the Japan Society for Promotion of Science (R.H.).


chemokine (C-X-C motif) ligand 14
breast and kidney-expressed chemokine
enzyme-linked immunosorbent assay


View Abstract