**1. Introduction**

Human exposure to air pollution is among major health problems. Some pollutants can be easily monitored, doses obtained by individuals may be calculated and the results of such exposure may be predicted if the "dose–e ffect" dependencies are known. These dependencies are not known for most bioaerosols. An atmospheric bioaerosol is omnipresent part of atmospheric aerosols, accounting for up to 95% of their total number or up to 80% of their mass [1–4]. The peculiarity of atmospheric bioaerosols is, besides the usual e ffects of aerosols on atmospheric processes and climate [5–7], the ability to cause or provoke various infectious or non-infectious diseases in humans [5,8–12]. The most important component of atmospheric bioaerosols is microorganisms. Since microorganisms are usually hazardous to humans, it is important to be able to assess what this hazard is in each air sample. It is also important to be able to track the change in this hazard in the air of controlled points over time. Modern molecular biological methods can quite quickly reveal the biological diversity of microorganisms present in a sample [13–27], but for the vast majority of known microorganisms, their danger to humans has not been studied. Therefore, an approach that is based on growth, morphological and biochemical properties of culturable microorganisms (even if a microorganism has never been isolated from the natural environment early or most of its properties have not been studied) may be very useful.

We have previously proposed such an approach for determining the hazard of bacteria and yeasts found in atmospheric bioaerosols for humans and demonstrated its capabilities using the available experimental data as an example [28]. In the framework of the developed approach, a dimensionless integral index of the hazard to humans of microorganisms isolated from atmospheric bioaerosols was constructed, which is the product of four lower-level dimensionless integral indices characterizing the complex properties of the microorganism. One of them is an integral index of the concentration of culturable bacteria in an aerosol sample. Other components of the integral index of the hazard of microorganisms to humans are [28]:


Obviously, the more culturable microorganisms (identical or different) are in the air, the more dangerous they are to humans. The value normalized to the maximum number of culturable microorganisms in all samples studied (or in one sample; in this case, of course, this index is equal to 1) represents the first index [28]. The second index assesses the hazard (pathogenicity or conditional pathogenicity) of microorganisms to humans. Its normalized value reaches 1 for pathogens and is strictly equal to zero for completely non-pathogenic microorganisms. This index is based on the morphological and biochemical properties of microorganisms and allows us to predict their individual or collective pathogenicity in the range of 0 to 1. The third normalized index assesses the resistance of microbial cells to adverse environmental factors. It considers their growth, morphological and biochemical properties. The more resistant the microorganism is in the external environment, the greater the likelihood of it entering the human body and maintaining the ability to trigger negative reactions. Finally, the fourth index is determined by the resistance of microorganisms to the action of antibiotics or other drugs. The higher this resistance, the more difficult it is to overcome the negative effects of microorganisms in the body and, consequently, the higher the hazard of such a microorganism. The approach proposed in [28] was illustrated using experimental data obtained in 2006–2008.

Changes occurring in nature (climate change, atmospheric processes changes, changes of habitats areas of animals, insects and vegetation, changes in water systems [29–34]) should also be manifested in changes in the abundance and biodiversity of bioaerosols in the atmosphere.

Western Siberia is a region in which global climatic changes are clearly manifested: permafrost thawing, decrease in snow cover time, temperature increase, powerful greenhouse gas emissions [35–45]. At the same time, changes in the state of health of the population of the region caused by these climatic changes are poorly studied (see, for example, [37,46,47]). The method we developed allows us to assess the change in the hazard to humans of cultivated microorganisms located in atmospheric aerosols, and as a result, the influence of this factor, which undoubtedly affects the health of the population of the region. Thus, the aim of this work is to obtain new data on the hazard to humans in the context of ongoing climate change, both for individual microorganisms (only bacteria and yeasts in this paper), which are present in atmospheric bioaerosol, as well as their pool, and compare the results with the results obtained previously. This study includes two on-ground sampling sites where a person breathes directly and where the presence of microorganisms is largely determined by local sources of bioaerosols. In addition, microorganisms in the atmosphere at altitudes of up to 7 km were studied and its bioaerosol composition was largely determined by bioaerosols remote sources.

#### **2. Materials and Methods**

To ensure the comparability of the results obtained in 2006–2008 and in 2012–2016, materials and methods used were mainly the same as described in [28]. These materials and methods which were used in 2006–2008 and in 2012–2016 are described in details below, and their differences are highlighted in special Section 2.4.

#### *2.1. Atmospheric Air Sampling*

Sampling of atmospheric air was carried out at three points in the studied region: on the site of the FBRI SRC VB "Vector" of Rospotrebnadzor (Vector), 4 times a day (the sampling starts at 10:00, 16:00, 22:00 and 4:00 on the next day) in the middle of the month; in the Klyuchi village, once a season for 7 consecutive days (usually the sampling starts at 8:00–9:00); about 50 km south of Novosibirsk using the "Optik-E" laboratory (Aircraft, Figures 1 and 2) on one of the last days of each month. A map showing the position of two sampling points and a typical flight trajectory of a laboratory airplane is shown in Figure 3. The "Optic-E" laboratory mounted on an Antonov-30 aircraft (Figure 2, see also [48–51]) for 2006–2008 session includes device for bioaerosol sampling and additional devices for registration of physical and chemical aerosol characteristics, meteorological conditions, etc. [48,52]. Isokinetic air sampling was performed through the special inlet outside the cabin (inserts in Figures 2 and 3) [48,53] at cruising speed of aircraft approximately 360 km/hour. The operation of the air intake with reduced dynamic pressure is based on the principle of Venturi. Further, outboard air with a pressure equal to the air pressure inside the aircraft cabin was supplied to a stand with impingers (Figure 4), into which it was sampled for analysis of the presence and concentration of culturable microorganisms. It is obvious that both the device for outboard air intake and the tubes leading to the impinger are characterized by a certain percentage of losses of aerosol particles during their transportation. However, since both the intake air device and the length of the tubes are unchanged in all samples, these losses are the same for all samples. The samples were taken during daytime (usually the time of flight was in interval 12:00–15:00) over the Karakan pine forest (sees Figure 3) successively at eight altitudes: 7000, 5500, 4000, 3000, 2000, 1500, 1000, and 500 m [28]. To create sterile conditions, before each sampling or each flight, all incoming tubes were rinsed with ethanol.

**Figure 1.** Photo of Antonov-30 aircraft. Insert is the samplers' inlets.

**Figure 2.** Photo of Tupolev-134 aircraft. Insert is the samplers' Inlets.

**Figure 3.** The map of sampling sites and aircraft's flight typical trajectory.

Stainless steel impingers with a critical nozzle [28] (its analogue is described in [54]) were used for air sample collection. These devices (manufactured by JSC "Experimental-design bureau of biological precision engineering", Kirishi, Russia) maintain a constant flow rate at a pressure differential of more than 4 × 10<sup>4</sup> Pa of air through the device. An A-D1-04 pump (JSC "Kot", St. Petersburg, Russia) was used to pump air samples through the impinger. Fifty milliliters of noncolored Hanks' solution (SIGMA) was used as a sorbing fluid. Above-ground samples were taken at the flow rate of 50 ± 5 L/min for 30 minutes and altitude samples for 5 or 10 minutes (10-minute sampling was used during 2006–2008 and 5-minute sampling was used during 2012–2016) at each altitude. The retention efficiency of this impinger for aerosols of more than 0.3 μm exceeds 80% making up a constant value of 90% ± 15% for particles with a diameter of more than 2 μm.

**Figure 4.** Photo of two impingers (red circles) at a sampling stand of microorganisms located in atmospheric aerosol.

#### *2.2. Culturable Microorganisms' Concentration*

The concentrations of culturable microorganisms were determined by standard microbiological methods. Samples were seeded onto Petri dishes containing agarized media. LB [55] was used to detect saprophyte bacteria; depleted LB medium (diluted 1:10) was used to isolate microorganisms inhibited by the excess of organic substances, starch–ammoniac medium [56] was used to detect actinomyces; soil agar was used for soil microorganisms, and Sabouraud medium [56] was used for lower fungi and yeast. Successive sample dilutions were prepared when necessary. The seedings were incubated in a thermostat at a temperature of 28–30 ◦C for 3–14 days. Some isolates were additionally incubated at a temperature of 6–10 ◦C in 2012–2016. Phase contrast light microscopy was used for the study of morphological characteristics of bacteria (live cells and fixed Gram-stained ones too) and its colonies. Taxonomic groups the detected microorganisms referred to were determined according to [57–59]. Nucleotide sequences of PCR products corresponding to the fragments of 16S rDNA gene was performed for some bacteria [60,61]. The numbers of culturable microorganisms in samples were calculated according to standard methods [62]; the number of microorganisms was averaged over 3–4 parallels of samples 4–5 seeded on different media.

#### *2.3. Microorganisms' Biochemical and Morphological Characteristics*

#### 2.3.1. Pathogenic Properties of the Isolates

Isolated microorganisms were tested for the presence of the following signs of pathogenicity:


#### 2.3.2. Growth Characteristics of Bacteria at Increased Salt Concentration

The resistance of the studied microorganisms to high salt concentrations was determined when they were sown on a complete nutrient medium with the addition of NaCl at concentrations of 1%, 5% or 10%. After incubation at the optimum temperature, the range of resistance of the tested microorganism to the concentration of salts in the nutrient medium was established by the nature of growth or its absence.

#### 2.3.3. The Determination of Enzymatic Activity of Isolated Bacteria

Isolated microorganisms were tested for the presence of the following signs of enzymatic activities:

	- 1. seeding cultures with a stroke on yolk nutrient agar, for the preparation of which, under aseptic conditions, in yolk of molten and cooled to 50–55 ◦C fish-peptone agar (FPA) medium, yolk from a chicken egg is introduced; then medium is thoroughly mixed and poured into Petri dishes; the studied culture is streaked, incubated at the required temperature for 24–48 hours and the result is taken into account; when observed in an oblique light, lipase production was judged by the formation of a pearly shiny hydrolysis zone on agar around grown colonies; lecithinase (phospholipase) hydrolyzes the yolk lecithin; as a result of the precipitation reaction, a turbid whitish zone forms around the lecithin-fermenting colonies [57];
	- 2. plating cultures on a complete LB or RPA medium with 1% Tween-20 or Tween-40 and 0.01% CaCl2 as a substrate; sown cultures were incubated in a thermostat for 3–4 days and the result of the presence or absence of hydrolysis zones was determined [57];

of the reaction mixture [59]; enzyme activity was determined within 3 hours of incubation by absorption on Uniplan apparatus (Russia) with a color filter at the wavelength of 450 nm.


#### 2.3.4. Microorganisms' Antibiotic Resistance

The sensitivity of microorganisms to antibiotics was determined by diffusion using the discs method [57]; the concentration of antibiotics in the discs used was: ampicillin (10 μg/disk), neomycin (30 μg/disk), benzyl-penicillin (100 U/disk), levomycetin (30 μg/disk), carbenicillin (100 μg/disk), canamycin (30 μg/disk), oleandomycin (15 μg/disk), rifampicin (5 μg/disk), streptomycin (30 μg/disk), polymxin (300 U), erythromycin (15 μg/disk), lincomycin (15 μg/disk), oxacillin (10 μg/disk), gentamycin (10 μg/disk), tetracycline (30 μg/disk), vancomycin (30 μg/disk), amikacin (30 μg/disk), netilmycin (30 μg/disk), monomycin (30 μg/disk). Note that lists of antibiotics used were not the same for 2006–2008 and for 2012–2016.

#### *2.4. Changes in the Methods Used in the 2006–2008 and 2012–2016 Studies*


#### *2.5. Data Analysis and Statistics*

The initial data for all measured values for which lower-level integral indices were calculated using the formulas published in [28] and these integral indices themselves are given in Supplementary Materials (Tables S1–S12 and List of description). The average values of the integral indices and their standard deviations from the mean are given in these tables for each year of observation. Additionally, the maximum values of the corresponding integral for each year are given.

In Table S13, the integral index of hazard of bacteria and yeasts found in atmospheric bioaerosols for humans are calculated from the lower level integral indices for each of the isolates according to the procedure described in [28]. Furthermore, using the method of regression and ANOVA analysis (built-in Microsoft Excel software), a statistical analysis of the data was carried out. All analysis results were obtained at a reliability level of 95%.
