In-vivo measurement of the fluorescence spectrum of wild cochineal (Dactylopius opuntiae)

In this section, the methodology for the fluorescence spectra acquisition of the in-vivo cochineal in its natural ambient is described. The optical setup used is our own design that guarantees the detection of low levels of fluorescence present in this study. In addition, the wild cochineal reproduction model employed to ensure the existence of the samples is described. This study proposes a fluorescence standard based on a commonly used disk-shaped carmine colored pencil segment because of the non-availability of a commercial one in our laboratory.

Optical setup for fluorescence spectra measurements

The optical setup for detecting fluorescence spectra consists of (1) a power supply, (2) an excitation source, (3) a dichroic mirror, (4) a 10 × microscope objective lens, (5) a homemade sample-holder, (6) a mechanical positioning device with micrometric movements in x, y and z, (7) a 5 × microscope objective lens, (8) multi-mode optical fiber, (9) a miniature fiber optic spectrometer and (10) computer equipment. The previous elements were arranged as shown in Fig. 5.

The power supply (M30-TP305E, Shanghai MCP Corp.) provides a constant voltage of 3.5 V to the excitation source. The excitation source is a commercial laser pointer (KG-303-6) emitting at a wavelength of 532 nm and nominal power of 1 W. This wavelength is found within the absorption band of carminic acid diluted in methanol¹⁹, and as was demonstrated by Cárdenas²³, this wavelength allows the excitation of the fluorescence of natural carminic acid from wild cochineal in-vivo. The dichroic mirror (DMLP550, Thorlabs, Inc.) is tilted at an angle of 45° with respect to the direction of laser beam propagation allowing the excitation light to be reflected towards the aperture of the 10 × microscope objective lens (M-10X, Newport, Co.) which in turn focuses this radiation on a plane where the sample is placed. The specimen to be studied is fixed on the sample holder which in turn is magnetically coupled to the mechanical positioning device (M-900, Newport, Co.), that together with a micrometric base (High-Performance Linear Stage, Newport, Co.) allows the cochineal to be placed at the focal point of the excitation beam. Both the fraction of the fluorescent light emitted isotropically by the cochineal and the fraction of the light reflected diffusely by this insect enter the 10 × objective lens and leave it as a collimated beam towards the dichroic mirror. The first fraction of light is transmitted approximately 85% to the second microscope objective lens (M-5X, Newport, Co.) while the second one is strongly blocked by the dichroic mirror. A 400 µm diameter optical fiber (QP400-2-SR, Ocean Optics, Inc.) is placed at the focal distance from this latter microscope objective to collect the fluorescence radiation and transmit it to the input port of the miniature fiber optic spectrometer (Exemplar, B&W Tek, LLC.) which allows for the spectral decomposition of the fluorescence radiation for its later processing. This spectrometer is connected through a USB port to a desktop computer where the BWSpec (B&W Tek, LLC.) software is installed, provided by the same manufacturer of this device, by means of which the acquisition of the spectra is carried out with an integration time of 200 ms. This software allows the user to save the spectra in plain text files (.txt) for further processing of the acquired information.

The use of the dichroic mirror in the optical setup allows for adequate discrimination of the excitation and emission wavelengths. It shows about a 90% reflection in the range of 380–535 nm, the range in which the wavelength of the excitation beam is found (λexc = 532 nm), whereas for wavelengths in the range of 565–800 nm it presents an 85% transmission. We expect that the in-vivo cochineal fluorescence emission should be mainly located in this last spectral region, based on the findings reported in previous research works for solutions of carminic acid in methanol^19,20,23.

Carminic acid fluorescence standard

As a part of this research, we proposed the use of a segment from a carmine colored pencil (Ekuz Carmín, AZOR) as a fluorescence standard due to the absence of a commercial one in our laboratory. This homemade fluorescence standard was used to calibrate the spectra acquisition optical setup. The standard was obtained by cutting a 5 mm segment of the carmine colored pencil with a razor blade (Stainless blade, Dorco, Co.) to achieve a finer cut. Then, this segment was situated inside a plastic piece in an annulus-shape to provide protection and handling during fluorescence measurements. Finally, its face of greater diameter was polished with fine grit sandpaper (B-99 1000, Fandeli, México) to get a smooth plane surface. The opposite face was adhered to the sample-holder while the polished one was placed perpendicular to the optical axis of the 10 × objective lens so that the excitation light impinges on it allowing the optimization of the fluorescence signal recorded by the optical setup.

In order to test the temporal stability of the carminic acid fluorescence standard, a continuous acquisition of fluorescence spectra was performed by using the optical setup previously shown in Fig. 5. To do so, the excitation light source was turned on 30 min before the measurements. Then, the spectra acquisition was done one after the other by allowing the excitation light hits on the standard to detect its fluorescence, immediately blocking this radiation while its spectrum was recorded, and so on until 10 spectra were recorded.

Biological samples and reproduction model

The cochineal samples used in this research come from a cladode infected with this pest donated by farmers from the Sociedad Cooperativa Productora Agropecuaria de Nopal Tlanalapa S.C. de R. L. de C.V., from the state of Hidalgo, Mexico.

For its survival, this pest was reproduced in the Biomedical Optics Laboratory of the Polytechnic University of Tulancingo, according to the reproduction model illustrated in Fig. 6, in a small-implemented orchard of plastic pots. In these pots were planted healthy adult cladodes of approximate dimensions of 40 × 20 × 2 cm; these cladodes were intentionally infected with this pest.

As observed, in a plastic pot with a diameter approximately of 20 cm, two healthy cladodes were planted with the cladode infected with the cochineal pest placed between them. The spread of the pest is carried out through the points of contact between cladodes, which in this case were the thorns of each specimen, the separation distance between them was 1.5–3 cm approximately. Since the first stage of cochineal pest is mobile, during migration to other parts of the cladode, some cochineals fell to the ground infecting it from the base.

If we consider that in literature it has been reported that during oviposition females deposit about 419 eggs⁴ and we note that in our reproduction model both sides of the infected cladode have about 30 colonies conformed of 1–4 adult cochineal per side, then it is assumed that the spread of the pest towards the healthy cladodes with an infected cladode is enough to contaminate. This was visually verified by observing that during the course of two days both cladodes were already contaminated. In other variants of the reproduction model it was verified that a cladode with the pest is enough to infect four to five adult cladodes approximately of similar dimensions to those used in this study. This reproduction procedure allows us to have several groups of pots where the cochineals are in different maturation stages. The reproduction model proposed in this work is similar to the previously reported reproduction methods^14,15,28,29 where the planting of the cladodes is also done in pots, however, the pest infestation is carried out in a different way from the one proposed in this work.

In order for the pest to maintain an appropriate lifecycle, the pots are taken out the laboratory to expose the cladodes to sunlight during the day, taking care to avoid the rain and predators that could put the cochineals at risk. In the evening, approximately at 6 o’clock, the pots are returned to their resting place in the laboratory for their care, conservation and adequate control for this study. To be specific, in this study a photoperiod of 9 ± 1 h of direct solar radiation was stablished, to later leave the pots inside the laboratory in absolute darkness, until the next day. This photoperiod was applied for 3 months, for covering the life cycle of the cochineals (90 days approx.). A script development in MATLAB R2014a (The MathWorks, Inc.) software was performed to process the solar radiation detected by the UPT weather station (Vantage Pro2, Davis Instruments Corp.), as shown in Fig. 7, during the last five days of photoperiod from June 01st to 05th, which correspond also to the end of life cycle of the cochineal.

Note that the photoperiod was established in the time range of 09:00–18:00 h, (0–540 min in graphs), corresponding to the time in which the reproduction model was exposed to direct sunlight, so that at the end of the day was placed again inside the laboratory. Figure 7 shows the following data: date, time of maximum solar radiation, also in minutes, and the value of solar radiation [W/m²], corresponding to each graph of this figure.

The values of temperature and humidity, external (blue line) and internal (red line), recorded by the UPT weather station during the fluorescence spectra acquisition period for each set presented in this work, were processed in a script developed in MATLAB R2014a (The MathWorks, Inc.) software. In Fig. 8 are shown temperature and humidity as a function of time in a temporal range of 35 min, approximately.

The measurement periods of each cochineal set, comprise around 35 min, where most of the cases were performed from 18:10 to 20:15 h. In this figure, data of temperature and humidity recorded by the UPT weather station were plotted, both external and internal with respect to the CIDETyP (Centro de Investigación, Desarrollo Tecnológico, Transferencia de Tecnología y Posgrado) building where the Biomedical Optics Laboratory is located. In Table 3 the temperature and humidity related to six replicates of our experiment are shown, since the cochineal life cycle (approximately 90 days) to the spectral measurement of the cochineal sets (set 0–5). These data are arranged by stage, date, temperature and humidity. These two lasts are in turn divided into the information collected from an online weather service page (WeatherOnline Online Services, www.weatheronline.mx) and by the UPT weather station (Vantage Pro2, Davis Instruments Corp., www.davisinstruments.com). The values, corresponding to temperature and humidity recorded by the UPT weather station, are the result of an average of the data plotted in Fig. 8.

Observe that, in Table 3 exist differences between the temperature and humidity provided by the online weather service page and the UPT weather station. In this regard, since the data from the UPT weather station contain local information of the environment where both the growth of the pest in the reproduction model and the experimental fluorescence measurements were made, these values of temperature and humidity have been adopted in this work.

Selection of the samples to study spectrally

For this study 3 live-female cochineals of different sizes were selected, of around 480 µm for the small cochineal, 540 µm for the medium cochineal and 790 µm for the largest cochineal. These sizes were determined with a script developed in MATLAB R2014a (The MathWorks, Inc.) software reported in the work of Cárdenas²³. We assumed that the cochineals are infesting the cladode, in other words feeding from the sap, given their morphological features and size observed and compared with those referred in the available literature⁴.

Sample preparation

In order to prepare the biological samples to be studied spectrally, a segment of approximately 6 mm long was cut from one of the infected cladodes planted in the laboratory. The three cochineals previously classified in different sizes were found in this portion as shown in Fig. 9a, where, according to their lifecycle, they are in the stage of nymph I (crawler) and II, that is, in their early growth stages with 23 days old. This cladode portion was fixed on the round base of a plastic cylinder with a glue stick (Lápiz adhesivo DIXON, 36 g). The other end of the round base was inserted inside of an aluminum ring to which two magnets were attached with adhesive (Kola Loka Brocha, 5 GRS) to the sides with 90° in respect to the other, which allows for the proper placement of the sample in the pinhole mount of the magnetic positioning device during fluorescence measurements. Figure 9b shows the lateral view of the sample-holder, where it can be seen that the plastic round base and the sample protrude from the aluminum ring. This device allows us to keep considerable distances and displacements between the sample and the 10 × objective microscope lens during fluorescence measurements as shown in Fig. 9c.

Measuring procedure of the fluorescence spectra

The in-vivo fluorescence spectra measurements from the cochineals were performed using the optical setup described at the beginning of this section under laboratory conditions of complete darkness, at a temperature range from 20 to 25 °C as well as a relative humidity range from 42 to 80%. Additionally, it was necessary to turn on the excitation laser 30 min before the measurements began in order to achieve a stable fluorescence signal. During this time the output of the laser was obstructed with a dark piece of cardboard to avoid photodegradation of the sample, which was already placed in the sample-holder. Subsequently, the sample was approached to the edge of a dark colored, 374 µm thin plastic sheet, located just at the focal distance of the objective lens. Knowing the thickness of this sheet allows for the sample to be placed at the focal distance of the microscope objective once the plastic sheet is removed by moving properly the sample with the micrometric mechanism along the z-axis. This guarantees that recorded spectra of the cochineal are carried out “correctly”, that is, when the excitation focal point is on the highest part of the surface of this insect.

Once in this location, 10 fluorescence spectra were acquired from each one of the cochineals selected and located in the cladode portion. Three displacements of the sample holder were performed on the z-axis, moving the sample away from and then back toward to the focus plane of the laser excitation beam. Finally, these fluorescence spectra were averaged using a script development in MATLAB R2014a (The MathWorks, Inc.) software, obtaining a smoothed average spectrum of each cochineal as well as the standard deviation between them.

Figure 9c shows a lateral view of the way in which the fluorescence measurements of the in-vivo cochineals were performed. This view, of the interaction area of the excitation beam with a cochineal, allow us to observe that the spot size of the excitation laser is wide enough to cover completely the body of the cochineal.

Statistical analysis

Ten fluorescence spectra were acquired from each cochineal of different size that was present in the biological samples. These spectra were processed with a script developed by the authors in MATLAB R2014a (The MathWorks, Inc.) software, obtaining a smoothed average spectrum of each cochineal as well as its standard deviation. Figures for comparative fluorescence spectra of cochineals of different sizes show average values of fluorescence intensity and its standard deviation (errors bars). Control of reference fluorescence spectra of the background medium (cladode) and a homemade fluorescence standard were taken in all the experiments. Experiments were replicated in five sets of biological samples containing cochineals of similar sizes as is shown in Fig. 10 (see Table 1, which is complementary to this figure below). Statistical analysis of the fluorescence intensity at four wavelengths namely 620, 640, 660 nm and 760 nm were carried out using ANOVA test for the three cochineal sizes (small, medium and large) in each of the five available samples. The statistical values were computed using the Minitab 16 (Minitab, LLC.) software, where the computed p values < 0.05 were considered to be statistically significant.

On the other hand, in order to support the comparison between fluorescence spectra mentioned in this work, we have processed and collected spectral information (see Table 2) from the different sets presented in Table 1. This information includes data about the spectral bandwidth, maximum emission wavelength and maximum fluorescence intensity of each cochineal classified in sizes of S, M and L that incorporates the repeatability sets presented above. These values correspond to those obtained with the MATLAB R2014a script which computes the fluorescence spectra averaged and shown in this work.

In order to support the hypothesis of this work by determining how the cochineal size modifies the intensity of in-vivo fluorescence spectra, we have developed a script of our ownership in MATLAB R2014a (The MathWorks, Inc.) software for achieving to fit this intensity as a depending function of the cochineal size in a linear model as observed in Fig. 3b–f. We have processed the ten spectra measured as indicated in the “measuring procedure of the fluorescence spectra” section, emitted by each cochineal from each set presented in Fig. 10, which the values of fluorescence intensity were selected at the wavelength of 640 nm. In Fig. 2 it can be seen that, at this wavelength, the spectra show remarkably visible differences in the fluorescence intensity emitted by the cochineal classified in sizes S, M and L, allowing to distinguish which spectrum corresponds to which size of cochineal. We use the MATLAB function “polyfit” to fit the averaged fluorescence intensity values in a linear model, obtaining five graphics that show the average fluorescence intensity as a function of the size cochineal. In order to address the quantitative point of view of this study, we have computed the Pearson correlation coefficient (r) for each set of fluorescence spectra by using the MATLAB function “corrcoef”. For statistical purposes, we have processed values of fluorescence intensity at three additionally wavelengths in 620 nm, 660 nm and 760 nm (data not shown), in order to obtain a linear fit and to calculate linear correlation coefficients.

Source: Ecology - nature.com