Overview of global literature

Using the strategy described above, a total of 231 relevant publications were collected from 2015 to 2022 After excluding six conference abstracts, two editorial materials, one letter, one early access publication, and one processing paper, a total of 220 publications were retrieved (including 164 articles and 56 reviews). Excluding 1 not in English language, 219 publications were finally identified (Fig. 1). Figure 2A shows that the trend of volume in the global papers fluctuates but has generally increased year on year, especially since 2019. The annual number of published papers has basically remained above 30, indicating that, as research and development of CRISPR-related technologies deepened, researchers became increasingly interested in delivery of CRISPR components through EVs. The same trend can be seen in the relative interest in the field. Figure 2B demonstrates that, during the past 7 years, the major part of such publications was produced by more than ten countries (or regions), with Mainland China having the most publications, followed by the USA, Germany, England, and Japan (Fig. 2C). Figure 2D showed that the annual production of articles from the top 10 countries/regions increased from 2 in 2015 to 55 in 2022. Mainland China started to formally publish articles focused on CRISPR related to EVs in 2015, which is earlier than four countries (or regions) in the top 10. However, the growth rate of its publications was faster than for most countries (or regions), except the USA and Germany. Overall, research on CRISPR related to EVs has increased in popularity among researchers in different regions and has reached a rapid developmental phase. The curve of the publication trend is shown in Fig. 2E, indicating that the number of publications in this field increased consistently from 2 in 2015 to nearly 300 by 2050 (R2 = 0.9220).

Fig. 1

Flowchart of the screening of the retrieved publications for this bibliometric analysis. Purple blocks represent the remaining sections after each cull, while orange blocks represent the culled literature. Numbers in parentheses are the corresponding number of publications. A total of 231 relevant publications were collected from 2015 to 2022 After excluding 6 conference abstracts, 2 editorial materials, 1 letter, 1 early-access publication, and 1 processing paper, a total of 220 publications were retrieved (including 164 articles and 56 reviews). Excluding 1 not in English language, 219 publications were finally identified

Fig. 2

General trend of related publication worldwide from 2015 to 2022. A The trend of RRI and number of publications over time. B The distribution of publications among countries. C The top 10 countries in the field and the proportion of different countries relative to China. D An alluvium plot of the number of publications in the top 10 countries over time. The area size represents the number of publications, while the slope of the line segment represents the growth rate of publications. E A linear regression plot based on curve fitting of the global publication volume from 2015 to 2022, predicting that there will be a total of 300 articles published by the middle of 2050

Assessment of publications from different countries and regions

Figure 3A shows that publications from the People’s Republic of China have the highest total number of citations (2857), followed by the USA with a total frequency of 2105, South Korea ranked third with 395, England with 299, and Japan with 270. Additionally, papers from South Korea had the highest average citation frequency (49.1). with the People’s Republic of China (34.4), Spain (29.3), and the USA (28.5) ranked second, third, and fourth, respectively (Fig. 3B). The statistics presented in Fig. 3C also indicate that the related publications from the People’s Republic China had the highest H-index (28), followed by the USA (27), England (10), anbd Germany (8). Although South Korea had the highest average citation frequency, it had the lowest H-index among the top 10 countries (5).

Fig. 3

A summary of the citation frequency of related publications worldwide to assess the quality of the publications. A The total citation frequency of each country in the top 10. B The average number of citations of published articles from different countries. C The enumeration and statistics of the highest H-index of the countries

Bibliometric analysis of global leading authors, countries, and institutions

Analysis of authors can reveal who is representative of a field and acts as the core force. According to Price’s law, authors who have published more than two papers are core authors in this field. According to the statistics of VOS, there are 141 core authors in this field, accounting for 8.4%, which meets Price’s law and indicate that a relatively stable cooperative group of authors has been formed in this field. Table 1 shows the top 6 in this field according to publications.

Table 1 Top 6 authors in the field of CRIPSR technology related to EVs

Besides the number of publications, the frequency of citations reflects the value of an author's research. We visualize the highly cited authors, countries, and institutions with the corresponding cooperation relationships in Fig. 4.

Fig. 4

Global leading authors, countries, and institutions. A Visualization diagram of authors with highly cited publications. B Visualization diagram of countries of highly cited publications. C Visualization diagram of institutions producing highly cited publications

Bibliometric analysis of citation bursts of authors, journals, and references

The cited author was analyzed and counted according to citation bursts from 2015 to 2022. Citation bursts represent researchers’ attention during a short period, reflecting whether an author’s research is close to the current hotspots. Figure 5A shows that the top 10 authors had strongest citation bursts. The strongest strength is 4.34, produced by Valadi H, followed by Thery Clotilde (4.02), Raposo G. (3.87), Kowal J. (3.52), and Li L. (3.21).

Fig. 5

Bibliometric analysis of citation bursts within authors, journals, and references. Red horizontal lines indicate the importance and attention of the authors, journals, and references in the field. A longer red line length indicates greater popularity for authors, journals, and references. A Analysis of citation bursts within authors. B Analysis of citation bursts within journals. C Analysis of citation bursts within references. All items ranked according to “Start year”

By analyzing the journals, we found that the emergence and development of open-source journals have contributed strongly to the advancement of this research field. There is widespread agreement that research should be freely available, although some scholars currently disagree with open-access journals. Figure 5B shows the top cited journals in this field during the past 7 years. The top cited journal is PLoS Genetics, showing the strongest bursts. This academic journal is focused on the field of genetics with a monthly publication cycle. The publication scope of this journal mainly covers genetics. The Journal of Biological Chemistry aims to elucidate the molecular and cellular foundations of biological processes through high-quality science, therefore focusing on papers with novel and important insights into mechanisms, rather than in a specific thematic area, making the JBC indeed a melting pot for interdisciplinary scientists. Moreover, EMBO Journal has published high-quality research with broad interest and impact on molecular and cellular biology, with a focus on physiological correlations and molecular mechanisms. Current Opinion in Cell Biology prefers to provide authoritative, comprehensive, and systematic reviews. Highly cited journals with strong bursts mainly involve fields such as genetics, biochemistry, molecular biology, immunology, etc.

The cited publications were further analyzed by first analyzing the top 10 cited papers in the field during 2015–2022 using VOSviewer (Fig. 5C). The first article with citation bursts was written by Colombo, attracting great interest after its publication. This article had a total strength of 4.05 and began to burst from 2016 to 2019. This review comprehensively summarizes the biogenesis, secretion, and subsequent fate of exosomes and other secreted extracellular vesicles [28]. Raposo showed a highest burst strength of 4.15 in 2018 [29]. This review describes the characterization of EVs and currently proposed mechanisms for their formation, targeting, and function. From 2015 to 2022, Kowal showed bursts for the longest period with a total strength of 2.55 [30]. This paper demonstrated the presence of exosomal and nonexosomal subpopulations within small EVs and proposes their differential separation by immuno-isolation using either CD63, CD81, or CD9, thus providing guidelines to define subtypes of EVs for future functional studies.

Bibliometric analysis of cooperation among authors, countries, and institutions

Co-citation means that an article is co-cited by different authors, or an author is cited multiple times. Co-citation can reflect the research direction and cross-development in academia. We counted all co-authored papers and thoroughly assessed the thematic relevance of each entry. We used VOSviewer to analyze all 141 authors who published over two papers. Figure 6A shows that the top 5 authors (where numbers in parentheses represent the total link strengths, as below): Wang [30], Zhang [27], He [25], Xu [24], and Duan [23]. Besides, 31 countries were included to analyze according to the same standard. The corresponding data are depicted and visualized in Fig. 6B. The top 5 countries were found to be the USA (66), England (34), Germany (31), the People’s Republic of China (23), and Canada (20). The top 5 institutions were (Fig. 6C): Guangzhou Med. Univ. (15), Zhengzhou Univ. (14), Chinese Univ. Hong Kong (11), Harvard Med. Sch. (11), and Shenzhen Univ. (10).

Fig. 6

Cooperation map among authorship, countries, and institutions. The size of each node represents the number of publications. The line between nodes represents the cooperative relationship, while the width of the lines indicates the degree of cooperation. Different colors represent different clusters. Nodes of the same color represent authors, countries, and institutions showing closer and more frequent cooperation. A Network visualization diagram of co-cited authors of the publications. B Network visualization diagram of cooperating countries among the publications. C Network visualization diagram of cooperating institutions among the publications

In summary, Fig. 6 reveals strong scientific collaboration among authors, institutions, and countries through rich links and nodes. The author cooperation map shows 78 nodes (Fig. 6A). Zhang H., Wang Y., Wang J., Zhang Y., and Zhao X.X. lead in terms of number of publications. The many scientific collaborations between researchers are represented by the dense linkages observed. Clustering effects were identified among the authors and demonstrated with colors (e.g., Wang Y., Zhang G.) and various nodes forming one cluster (Zhang Y.Y., Wang J.) as well as some nodes clustering into another one (Xu L.M., Xu X, Duan L.) and some nodes clustering into one cluster. Figure 6B reveals that the distribution of publications in the field of CRISPR-related EVs is not balanced among different countries. China and the USA account for many publications. They were also the closest research partners, and cooperate closely with many countries, including Australia, South Korea, England, Japan, Germany, etc. Excluding the influence of China and the USA, England, Japan, Australia, Spain, Switzerland, Germany, and other countries also cooperated with each other, but most of them not as closely as with China and the USA. These results confirm that the USA and China have a certain major position. Figure 6C indicates the general cooperation between institutions via nodes and connecting lines. The core institutions in this field include Harvard Medical School, Sun-Yat-Sen University, Nanjing Medical University, Fudan University, etc. Compared with Fig. 6A, the clusters among institutions are not obvious, albeit indicating an extensive, diverse, and abundant cooperation among institutions.

Interestingly, nodes of larger size may not have stronger connecting lines. This condition is particularly evident regarding the countries and institutions. This indicates that the research and paper output in this field is currently restricted by country or geographically. There is a clear internal tendency for researchers to favor working with nationals when collaborating. The core institutions that publish papers are not necessarily those that cooperate more closely with other institutions, which implies “clustering” in current research in this field.

Bibliometric analysis of co-cited journals and references

The co-citation network among journals can be seen to consist of roughly four clusters, shown in four colors in Fig. 7A. The top 3 cited journals are Nature (citations 419), Proc. Natl Acad. Sci. USA (citations 384), Nat. Commun. (citations 378). All of these are excellent journals in the JCR1 category.

Fig. 7

A Network visualization diagram of cooperating journals among the publications. B Network visualization diagram of co-cited publications. C Dual map overlay of journals related to CRISPR associated with EVs

Among these four clusters, except for the comprehensive journals, the red and blue clusters are mainly related to chemistry and materials, focusing on applications of different types of materials in various aspects. These journals are mainly cited to review and analyze existing studies in the field of materials science and provide previous theoretical and technical support for the development of new materials. The yellow and green journals are mainly in the field of medicine and translational medicine, focusing on specific applications of CRIPSR technology and EVs in the medical field or physiopathological mechanisms behind their therapeutic effects. Citation of publications in this field helps to explore CRISPR-related EVs to broaden understanding of this research area and provide theoretical guidance for the eventual translation of biomaterials and medical engineering.

The co-citation map reveals a co-citation network of highly co-cited literature. Figure 7B shows that this network can roughly be divided into three major clusters, shown in corresponding colors. The blue cluster represents literature in the field of CRISPR/Cas technology [31,32,33], while the green cluster is mainly related to the application of EVs in delivery systems [34,35,36]. The publications in the red cluster are mainly literature reviews on the characteristics of EVs [28, 29, 37,38,39,40]. Figure 7C shows a predominate citation path in orange, revealing that papers published in molecular/biology/genetics tend to cite papers from this same field.

Bibliometric analysis of keywords, active topics, and orientation in the future

The objective of keywords and co-occurrence analyses is to identify potential directions and hotspots of research. Such statistics can be used to monitor developments in scientific research. We screened 317 keywords that occurred ≥ 2 times in titles/abstracts within the 219 retrieved papers and visualize them in Fig. 8A. Hotspots in the area are represented by the frequency of these keywords. The keywords are shown in larger fonts and with larger nodes in the map, the more frequently they occurred. To provide a clearer picture of the specific keywords, high-frequency keywords with frequency above 15 are also presented in Table 2. According to Fig. 8A and Table 2, high-frequency keywords such as expression, cells, and Cas9 are representative terms in this field.

Fig. 8

Bibliometric analysis of active topics, keywords, and orientation in the future. A Top keywords with the highest citation frequency based on CiteSpace. B Co-occurrence analysis of keywords. C Clustering analysis of the keyword network based on CiteSpace. D Timeline diagram of keywords with corresponding changes of cluster

Table 2 High-frequency keywords in the field of CRIPSR technology related to EVs

The co-occurrence analysis and clusters among keywords demonstrate the internal relationship among keywords. The keyword co-occurrence network view of the total 317 keywords is visualized in Fig. 8B. Larger nodes indicate more frequent appearance and greater representativeness of the domain hotspot; connecting lines represent the association strength, where a thicker line indicates that two keywords appear together in the same publication more often; the node color represents different clusters, i.e., research topics. The keywords identified were classified into seven clusters (Fig. 8C): cluster 0: exosome (red); cluster 1: nanovesicles (yellow); cluster 2: DNA (light green); cluster 3: gene editing (dark green); cluster 4: gene therapy (blue); cluster 5: cancer therapy (purple); and cluster 6: endometrial stromal cells (pink). The clusters indicate the most prominent topics in the field so far. In the “exosome” cluster, the primary keyword was “exosm”. For the “nanovesicles” cluster, the frequently used keyword was “drug-delivery”. For the “DNA” cluster, the main keyword was “microvesicles”. For the “gene editing” cluster, the dominantly used keyword was “extracellular vesicles”. For the “gene therapy” cluster, the frequently used keyword was “biogenesis”. In the “cancer therapy” cluster, the primary used keyword was “breast cancer”. In the “endometrial stromal cells”, the main used keyword was “disease”.

The clustering analysis of the keywords in terms of period allows to derive the baseline status of the research topics within the research field. To clearly identify the temporal patterns of inflection points and frontiers of disciplinary development, the keyword co-occurrence mapping can be arranged in a time series, thus revealing the distribution of research hotspots in each period. In this study, we selected the nodes as “Keywords”, set the “Slice Length” to 1, and set the “Selection Criteria” to “Top 50 per slice” in CiteSpace software, which meant we extracted the data of each time slice, ranking 50 to generate the keyword co-occurrence timeline (Fig. 8D). As seen in Fig. 7D, worldwide research on EVs based on CRISPR-related technologies started to proliferate in 2015 and evolved over a 7-year period. Combined with the qualitative analysis, this can be divided into two phases: the first phase is from 2015 to 2020, and as shown by the keyword clustering, most of the research at that time focused on diseases, disease treatment targets, and exploring the nature of EVs and CRISPR themselves. Since the secretion of EVs must be stimulated by signaling pathways and CRISPR is essentially a tool for gene editing, many keywords related to genetic laws, gene regulation, and targeting sites can be seen. This phase of research topics is relatively more focused. The second phase is from 2020 to 2022, when keywords begin to appear decentralized, with expanding research themes and a continuous increase in research hotspots. These include the selection of specific delivery options, in-depth exploration of CRISPR and vesicle-related technologies, well-defined gene therapies, and the study and exploration of a range of biomaterials, mainly nanomaterials. This indicates that scholars from different backgrounds and directions are converging here, while the research is moving from mechanism exploration and analysis and technology validation to the field of translational medicine, step by step.